Summary:
Make the test more portable and do not rely on the pre-bundled object
file.
Reviewers: Hahnfeld, hfinkel, jdoerfert
Subscribers: caomhin, kkwli0, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D66296
llvm-svn: 369015
This fixes a regression from r365860: As that commit message
states, there are 3 valid states targeted by the combination of
-f(no-)omit-frame-pointer and -m(no-)omit-leaf-frame-pointer.
After r365860 it's impossible to get from state 10 (omit just
leaf frame pointers) to state 11 (omit all frame pointers)
in a single command line without getting a warning.
This change restores that functionality.
Fixes PR42966.
Differential Revision: https://reviews.llvm.org/D66142
llvm-svn: 368728
Support -march=tigerlake for x86.
Compare with Icelake Client, It include 4 more new features ,they are
avx512vp2intersect, movdiri, movdir64b, shstk.
Patch by Xiang Zhang (xiangzhangllvm)
Differential Revision: https://reviews.llvm.org/D65840
llvm-svn: 368543
The default behavior of Clang's indirect function call checker will replace
the address of each CFI-checked function in the output file's symbol table
with the address of a jump table entry which will pass CFI checks. We refer
to this as making the jump table `canonical`. This property allows code that
was not compiled with ``-fsanitize=cfi-icall`` to take a CFI-valid address
of a function, but it comes with a couple of caveats that are especially
relevant for users of cross-DSO CFI:
- There is a performance and code size overhead associated with each
exported function, because each such function must have an associated
jump table entry, which must be emitted even in the common case where the
function is never address-taken anywhere in the program, and must be used
even for direct calls between DSOs, in addition to the PLT overhead.
- There is no good way to take a CFI-valid address of a function written in
assembly or a language not supported by Clang. The reason is that the code
generator would need to insert a jump table in order to form a CFI-valid
address for assembly functions, but there is no way in general for the
code generator to determine the language of the function. This may be
possible with LTO in the intra-DSO case, but in the cross-DSO case the only
information available is the function declaration. One possible solution
is to add a C wrapper for each assembly function, but these wrappers can
present a significant maintenance burden for heavy users of assembly in
addition to adding runtime overhead.
For these reasons, we provide the option of making the jump table non-canonical
with the flag ``-fno-sanitize-cfi-canonical-jump-tables``. When the jump
table is made non-canonical, symbol table entries point directly to the
function body. Any instances of a function's address being taken in C will
be replaced with a jump table address.
This scheme does have its own caveats, however. It does end up breaking
function address equality more aggressively than the default behavior,
especially in cross-DSO mode which normally preserves function address
equality entirely.
Furthermore, it is occasionally necessary for code not compiled with
``-fsanitize=cfi-icall`` to take a function address that is valid
for CFI. For example, this is necessary when a function's address
is taken by assembly code and then called by CFI-checking C code. The
``__attribute__((cfi_jump_table_canonical))`` attribute may be used to make
the jump table entry of a specific function canonical so that the external
code will end up taking a address for the function that will pass CFI checks.
Fixes PR41972.
Differential Revision: https://reviews.llvm.org/D65629
llvm-svn: 368495
Fixes PR16786
Currently, library paths specified by LIBRARY_PATH are placed after inputs: `inputs LIBRARY_PATH stdlib`
In gcc, the order is: `LIBRARY_PATH inputs stdlib` if not cross compiling.
(On Darwin targets, isCrossCompiling() always returns false.)
This patch changes the behavior to match gcc.
Reviewed By: hfinkel
Differential Revision: https://reviews.llvm.org/D65880
llvm-svn: 368245
Globals are instrumented by adding a pointer tag to their symbol values
and emitting metadata into a special section that allows the runtime to tag
their memory when the library is loaded.
Due to order of initialization issues explained in more detail in the comments,
shadow initialization cannot happen during regular global initialization.
Instead, the location of the global section is marked using an ELF note,
and we require libc support for calling a function provided by the HWASAN
runtime when libraries are loaded and unloaded.
Based on ideas discussed with @evgeny777 in D56672.
Differential Revision: https://reviews.llvm.org/D65770
llvm-svn: 368102
There are times when we wish to explicitly control the C++ standard
library search paths used by the driver. For example, when we're
building against the Android NDK, we might want to use the NDK's C++
headers (which have a custom inline namespace) even if we have C++
headers installed next to the driver. We might also be building against
a non-standard directory layout and wanting to specify the C++ standard
library include directories explicitly.
We could accomplish this by passing -nostdinc++ and adding an explicit
-isystem for our custom search directories. However, users of our
toolchain may themselves want to use -nostdinc++ and a custom C++ search
path (libc++'s build does this, for example), and our added -isystem
won't respect the -nostdinc++, leading to multiple C++ header
directories on the search path, which causes build failures.
Add a new driver option -stdlib++-isystem to support this use case.
Passing this option suppresses adding the default C++ library include
paths in the driver, and it also respects -nostdinc++ to allow users to
still override the C++ library paths themselves.
It's a bit unfortunate that we end up with both -stdlib++-isystem and
-cxx-isystem, but their semantics differ significantly. -cxx-isystem is
unaffected by -nostdinc++ and is added to the end of the search path
(which is not appropriate for C++ standard library headers, since they
often #include_next into other system headers), while -stdlib++-isystem
respects -nostdinc++, is added to the beginning of the search path, and
suppresses the default C++ library include paths.
Differential Revision: https://reviews.llvm.org/D64089
llvm-svn: 367982
On a musl-based Linux distribution, stdalign.h stdarg.h stdbool.h stddef.h stdint.h stdnoreturn.h are expected to be provided by musl (/usr/include), instead of RESOURCE_DIR/include.
Reorder RESOURCE_DIR/include to fix the search order problem.
(Currently musl doesn't provide stdatomic.h. stdatomic.h is still found in RESOURCE_DIR/include.)
gcc on musl has a similar search order:
```
/usr/lib/gcc/x86_64-alpine-linux-musl/8.3.0/../../../../include/c++/8.3.0
/usr/lib/gcc/x86_64-alpine-linux-musl/8.3.0/../../../../include/c++/8.3.0/x86_64-alpine-linux-musl
/usr/lib/gcc/x86_64-alpine-linux-musl/8.3.0/../../../../include/c++/8.3.0/backward
/usr/local/include
/usr/include/fortify
/usr/include
/usr/lib/gcc/x86_64-alpine-linux-musl/8.3.0/include
```
This is different from a glibc-based distribution where RESOURCE_DIR/include is placed before SYSROOT/usr/include.
According to the maintainer of musl:
> musl does not support use/mixing of compiler-provided std headers with its headers, and intentionally has no mechanism for communicating with such headers as to which types have already been defined or still need to be defined. If the current include order, with clang's headers before the libc ones, works in some situations, it's only by accident.
Reviewed by: phosek
Differential Revision: https://reviews.llvm.org/D65699
llvm-svn: 367981
Builtins-*-sunos :: compiler_rt_logbf_test.c currently FAILs on Solaris, both SPARC and
x86, 32 and 64-bit.
It turned out that this is due to different behaviour of logb depending on the C
standard compiled for, as documented on logb(3M):
RETURN VALUES
Upon successful completion, these functions return the exponent of x.
If x is subnormal:
o For SUSv3-conforming applications compiled with the c99 com-
piler driver (see standards(7)), the exponent of x as if x
were normalized is returned.
o Otherwise, if compiled with the cc compiler driver, -1022,
-126, and -16382 are returned for logb(), logbf(), and
logbl(), respectively.
Studio c99 and gcc control this by linking with the appropriate version of values-xpg[46].o, but clang uses neither of those.
The following patch fixes this by following what gcc does, as corrected some time ago in
Fix use of Solaris values-Xc.o (PR target/40411)
https://gcc.gnu.org/ml/gcc-patches/2018-01/msg02350.html and
https://gcc.gnu.org/ml/gcc-patches/2018-01/msg02384.html.
Tested on x86_64-pc-solaris2.11, sparcv9-sun-solaris2.11, and x86_64-pc-linux-gnu.
Differential Revision: https://reviews.llvm.org/D64793
llvm-svn: 367866
These "sanitizers" are hardened ABIs that are wholly orthogonal
to the SanitizerCoverage instrumentation.
Differential Revision: https://reviews.llvm.org/D65715
llvm-svn: 367799
This change introduces a pair of -fsanitize-link-runtime and
-fno-sanitize-link-runtime flags which can be used to control linking of
sanitizer runtimes. This is useful in certain environments like kernels
where existing runtime libraries cannot be used.
Differential Revision: https://reviews.llvm.org/D65029
llvm-svn: 367794
Summary:
question-mark is not a BRE special character.
POSIX.1-2017 XBD Section 9.3.2 indicates that the interpretation of `\?`
as used by rC366282 is undefined. This patch uses an ERE instead.
Reviewers: rnk, daltenty, xingxue, jasonliu
Reviewed By: rnk
Subscribers: cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D65668
llvm-svn: 367709
UBSan-Standalone-x86_64 :: TestCases/TypeCheck/Function/function.cpp currently
FAILs on Solaris/x86_64:
clang-9: error: unsupported option '-fsanitize=function' for target 'x86_64-pc-solaris2.11'
AFAICS, there's nothing more to do then enable that sanitizer in the driver (for x86 only),
which is what this patch does, together with updating another testcase.
Tested on x86_64-pc-solaris2.11.
Differential Revision: https://reviews.llvm.org/D64488
llvm-svn: 367351
The as-options.s test writes to the build tree as of r367165. Some build systems configure this to be readonly, so this fails. Explicitly write to the output tree using `%t` to avoid this.
llvm-svn: 367253
Same kind of fix as in r367176, but for "RUN on line 76"
this time.
I'll ask for a post-commit review, to ensure this
matches the intention with the test added in r367165.
But I think this at least will make the buildbots a
little bit happier.
llvm-svn: 367182
This morally relands r365703 (and r365714), originally reviewed at
https://reviews.llvm.org/D64527, but with a different implementation.
Relanding the same approach with a fix for the revert reason got a bit
involved (see https://reviews.llvm.org/D65108) so use a simpler approach
with a more localized implementation (that in return duplicates code
a bit more).
This approach also doesn't validate flags for the integrated assembler
if the assembler step doesn't run.
Fixes PR42066.
Differential Revision: https://reviews.llvm.org/D65233
llvm-svn: 367165
This restores the use of `rm` instead of the non-portable `ln -n`. Such
use being the status quo for the 12-month period between rC334972 and
rC365414.
llvm-svn: 367147
Summary:
r366702 added a set of new clang-cl -- specific openmp flags together with tests.
The way the newly added tests work is problematic: consider for example this
asertion:
```
// RUN: %clang_cl --target=x86_64-windows-msvc /openmp -### -- %s 2>&1 | FileCheck --check-prefix=CHECK-CC1-OPENMP %s
...
// CHECK-CC1-OPENMP: "-fopenmp"
```
It asserts that an `/openmp` flag should expand into `-fopenmp`. This however
depends on the default value of Clang's CLANG_DEFAULT_OPENMP_RUNTIME value.
Indeed, the code that adds `-fopenmp` to the output only does it if the default
runtime is `libomp` or `libiomp5`, not when it is `libgomp`.
I've updated the tests to not depend on the default value of this setting by
specifying the runtime to use explicitly in each assertion.
Reviewers: jdoerfert
Subscribers: cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D65272
llvm-svn: 367012
Rename lang mode flag to -cl-std=clc++/-cl-std=CLC++
or -std=clc++/-std=CLC++.
This aligns with OpenCL C conversion and removes ambiguity
with OpenCL C++.
Differential Revision: https://reviews.llvm.org/D65102
llvm-svn: 367008
Summary:
Move `-ftime-trace-granularity` option to frontend options. Without patch
this option is showed up in the help for any tool that links libSupport.
Reviewers: sammccall
Subscribers: hiraditya, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D65202
llvm-svn: 366911
with '-mframe-pointer'
After D56351 and D64294, frame pointer handling is migrated to tri-state
(all, non-leaf, none) in clang driver and on the function attribute.
This patch makes the frame pointer handling cc1 option tri-state.
Reviewers: chandlerc, rnk, t.p.northover, MaskRay
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D56353
llvm-svn: 366645
The RISC-V hard float calling convention requires the frontend to:
* Detect cases where, once "flattened", a struct can be passed using
int+fp or fp+fp registers under the hard float ABI and coerce to the
appropriate type(s)
* Track usage of GPRs and FPRs in order to gate the above, and to
determine when signext/zeroext attributes must be added to integer
scalars
This patch attempts to do this in compliance with the documented ABI,
and uses ABIArgInfo::CoerceAndExpand in order to do this. @rjmccall, as
author of that code I've tagged you as reviewer for initial feedback on
my usage.
Note that a previous version of the ABI indicated that when passing an
int+fp struct using a GPR+FPR, the int would need to be sign or
zero-extended appropriately. GCC never did this and the ABI was changed,
which makes life easier as ABIArgInfo::CoerceAndExpand can't currently
handle sign/zero-extension attributes.
Re-landed after backing out 366450 due to missed hunks.
Differential Revision: https://reviews.llvm.org/D60456
llvm-svn: 366480
Starting with Solaris 11.4 (which is now the required minimal version), Solaris does
support __cxa_atexit. This patch reflects that.
One might consider removing the affected tests altogether instead of inverting them,
as is done on other targets.
Besides, this lets two ASan tests PASS:
AddressSanitizer-i386-sunos :: TestCases/init-order-atexit.cc
AddressSanitizer-i386-sunos-dynamic :: TestCases/init-order-atexit.cc
Tested on x86_64-pc-solaris2.11 and sparcv9-sun-solaris2.11.
Differential Revision: https://reviews.llvm.org/D64491
llvm-svn: 366305
There was a slight typo in r364352 that ended up causing our backend to
complain on some x86 Android builds. This CL fixes that.
Differential Revision: https://reviews.llvm.org/D64781
llvm-svn: 366276
I noticed that clang currently passes --dynamic-linker to ld. This has been the case
since Solaris 11 support was added initially back in 2012 by David Chisnall (r150580).
I couldn't find any patch submission, let alone a justification, for this, and it seems
completely useless: --dynamic-linker is a gld compatibility form of the option, the
native option being -I. First of all, however, the dynamic linker passed is simply the
default, so there's no reason at all to specify it in the first place.
This patch removes passing the option and adjusts the affected testcase accordingly.
Tested on x86_64-pc-solaris2.11 and sparcv9-sun-solaris2.11.
Differential Revision: https://reviews.llvm.org/D64493
llvm-svn: 366202
i.e., recent 5745eccef54ddd3caca278d1d292a88b2281528b:
* Bump the function_type_mismatch handler version, as its signature has changed.
* The function_type_mismatch handler can return successfully now, so
SanitizerKind::Function must be AlwaysRecoverable (like for
SanitizerKind::Vptr).
* But the minimal runtime would still unconditionally treat a call to the
function_type_mismatch handler as failure, so disallow -fsanitize=function in
combination with -fsanitize-minimal-runtime (like it was already done for
-fsanitize=vptr).
* Add tests.
Differential Revision: https://reviews.llvm.org/D61479
llvm-svn: 366186
Summary:
This is a reland of r366146, adding in the previously missing '--'
flag that prevents filenames from being interpreted as flags.
Original description:
This adds a -fthinlto-index= option to clang-cl, which allows it to
be used to drive ThinLTO backend passes. This allows clang-cl to be
used for distributed ThinLTO.
Tags: #clang
llvm-svn: 366165
Summary:
This adds a -fthinlto-index= option to clang-cl, which allows it to
be used to drive ThinLTO backend passes. This allows clang-cl to be
used for distributed ThinLTO.
Reviewers: tejohnson, pcc, rnk
Subscribers: mehdi_amini, steven_wu, dexonsmith, arphaman, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D64458
llvm-svn: 366146
Summary:
Previously, passing -fthinlto-index= to clang required that bitcode
files be explicitly marked by -x ir. This change makes us detect files
with object file extensions as bitcode files when -fthinlto-index= is
present, so that explicitly marking them is no longer necessary.
Explicitly specifying -x ir is still accepted and continues to be part
of the test case to ensure we continue to support it.
Reviewers: tejohnson, rnk, pcc
Subscribers: mehdi_amini, steven_wu, dexonsmith, arphaman, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D64610
llvm-svn: 366127
Add "memtag" sanitizer that detects and mitigates stack memory issues
using armv8.5 Memory Tagging Extension.
It is similar in principle to HWASan, which is a software implementation
of the same idea, but there are enough differencies to warrant a new
sanitizer type IMHO. It is also expected to have very different
performance properties.
The new sanitizer does not have a runtime library (it may grow one
later, along with a "debugging" mode). Similar to SafeStack and
StackProtector, the instrumentation pass (in a follow up change) will be
inserted in all cases, but will only affect functions marked with the
new sanitize_memtag attribute.
Reviewers: pcc, hctim, vitalybuka, ostannard
Subscribers: srhines, mehdi_amini, javed.absar, kristof.beyls, hiraditya, cryptoad, steven_wu, dexonsmith, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D64169
llvm-svn: 366123
gcc PowerPC supports 3 representations of long double:
* -mlong-double-64
long double has the same representation of double but is mangled as `e`.
In clang, this is the default on AIX, FreeBSD and Linux musl.
* -mlong-double-128
2 possible 128-bit floating point representations:
+ -mabi=ibmlongdouble
IBM extended double format. Mangled as `g`
In clang, this is the default on Linux glibc.
+ -mabi=ieeelongdouble
IEEE 754 quadruple-precision format. Mangled as `u9__ieee128` (`U10__float128` before gcc 8.2)
This is currently unavailable.
This patch adds -mabi=ibmlongdouble and -mabi=ieeelongdouble, and thus
makes the IEEE 754 quadruple-precision long double available for
languages supported by clang.
Reviewed By: hfinkel
Differential Revision: https://reviews.llvm.org/D64283
llvm-svn: 366044
Using noexcept multilib with -fno-exceptions can lead to significant
space savings when statically linking libc++abi because we don't need
all the unwinding and demangling code.
When compiling with ASan, we already get a lot of overhead from the
instrumentation itself, when statically linking libc++abi, that overhead
is even larger.
Having the noexcept variant for ASan can help significantly, we've seen
more than 50% size reduction in our system image, which offsets the cost
of having to build another multilib.
Differential Revision: https://reviews.llvm.org/D64140
llvm-svn: 365994
The default implementation of getSupportedSanitizers isn't able to turn
on the vptr sanitizer, and thus, any platform that runs this test will
fail with the error:
clang: error: unsupported option '-fsanitize=vptr' for target '<target>'
Patch by James Nagurne!
llvm-svn: 365981
Summary:
This paves the way for using passive segments in pthread builds, which
will make separate memory files unnecessary. Mutable globals are also
necessary for the upcoming implementation of TLS.
Reviewers: aheejin, dschuff
Subscribers: sbc100, jgravelle-google, sunfish, jfb, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D64586
llvm-svn: 365935
This patch series adds support for the next-generation arch13
CPU architecture to the SystemZ backend.
This includes:
- Basic support for the new processor and its features.
- Support for low-level builtins mapped to new LLVM intrinsics.
- New high-level intrinsics in vecintrin.h.
- Indicate support by defining __VEC__ == 10303.
Note: No currently available Z system supports the arch13
architecture. Once new systems become available, the
official system name will be added as supported -march name.
llvm-svn: 365933
This patch makes the driver option -mlong-double-128 available for X86
and PowerPC. The CC1 option -mlong-double-128 is available on all targets
for users to test on unsupported targets.
On PowerPC, -mlong-double-128 uses the IBM extended double format
because we don't support -mabi=ieeelongdouble yet (D64283).
Reviewed By: rnk
Differential Revision: https://reviews.llvm.org/D64277
llvm-svn: 365866
Use a tri-state enum to represent shouldUseFramePointer() and
shouldUseLeafFramePointer().
This simplifies the logic and fixes PR9825:
-fno-omit-frame-pointer doesn't imply -mno-omit-leaf-frame-pointer.
and PR24003:
/Oy- /O2 should not omit leaf frame pointer: this matches MSVC x86-32.
(/Oy- is a no-op on MSVC x86-64.)
and:
when CC1 option -mdisable-fp-elim if absent, -momit-leaf-frame-pointer
can also be omitted.
The new behavior matches GCC:
-fomit-frame-pointer wins over -mno-omit-leaf-frame-pointer
-fno-omit-frame-pointer loses out to -momit-leaf-frame-pointer
The behavior makes lots of sense. We have 4 states:
- 00) leaf retained, non-leaf retained
- 01) leaf retained, non-leaf omitted (this is invalid)
- 10) leaf omitted, non-leaf retained (what -momit-leaf-frame-pointer was designed for)
- 11) leaf omitted, non-leaf omitted
"omit" options taking precedence over "no-omit" options is the only way
to make 3 valid states representable with -f(no-)?omit-frame-pointer and
-m(no-)?omit-leaf-pointer.
Reviewed By: ychen
Differential Revision: https://reviews.llvm.org/D64294
llvm-svn: 365860
This flag is analoguous to other flags like -nostdlib or -nolibc
and could be used to disable linking of profile runtime library.
This is useful in certain environments like kernel, where profile
instrumentation is still desirable, but we cannot use the standard
runtime library.
llvm-svn: 365808
clang currently warns when passing flags for the assembler (e.g.
-Wa,-mbig-obj) to an invocation that doesn't run the assembler (e.g.
-E).
At first sight, that makes sense -- the flag really is unused. But many
other flags don't have an effect if no assembler runs (e.g.
-fno-integrated-as, -ffunction-sections, and many others), and those
currently don't warn. So this seems more like a side effect of how
CollectArgsForIntegratedAssembler() is implemented than like an
intentional feature.
Since it's a bit inconvenient when debugging builds and adding -E,
always call CollectArgsForIntegratedAssembler() to make sure assembler
args always get claimed. Currently, this affects only these flags:
-mincremental-linker-compatible, -mimplicit-it= (on ARM), -Wa, -Xassembler
It does have the side effect that assembler options now need to be valid
even if -E is passed. Previously, `-Wa,-mbig-obj` would error for
non-coff output only if the assembler ran, now it always errors. This
too makes assembler flags more consistent with all the other flags and
seems like a progression.
Fixes PR42066.
Differential Revision: https://reviews.llvm.org/D64527
llvm-svn: 365703
D63793 removed float-divide-by-zero from the "undefined" set but it
failed to add it to getSupportedSanitizers(), thus the sanitizer is
rejected by the driver:
clang-9: error: unsupported option '-fsanitize=float-divide-by-zero' for target 'x86_64-unknown-linux-gnu'
Also, add SanitizerMask::FloatDivideByZero to a few other masks to make -fsanitize-trap, -fsanitize-recover, -fsanitize-minimal-runtime and -fsanitize-coverage work.
Reviewed By: rsmith, vitalybuka
Differential Revision: https://reviews.llvm.org/D64317
llvm-svn: 365587
This is a better fix for the problem fixed in r334972.
Also remove the rm'ing of the symlink destination that was there to
clean up the bots -- it's over a year later, bots should be happy now.
Differential Revision: https://reviews.llvm.org/D64301
llvm-svn: 365414
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.)
Differential Revision: https://reviews.llvm.org/D64253
llvm-svn: 365413
-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.
Reviewed By: rnk
Differential Revision: https://reviews.llvm.org/D64067
llvm-svn: 365412
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.
Reviewers: rnk, BillyONeal
Subscribers: cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D63793
llvm-svn: 365272
Summary:
LLVM issues a warning if passed unknown target features. Neither I nor
@asb noticed this until after https://reviews.llvm.org/D63498 landed.
This patch stops passing the (unknown) "save-restore" target feature to
the LLVM backend, but continues to emit a warning if a driver asks for
`-msave-restore`. The default of assuming `-mno-save-restore` (and
emitting no warnings) remains.
Reviewers: asb
Reviewed By: asb
Subscribers: rbar, johnrusso, simoncook, apazos, sabuasal, niosHD, kito-cheng, shiva0217, jrtc27, zzheng, edward-jones, rogfer01, MartinMosbeck, brucehoult, the_o, rkruppe, PkmX, jocewei, psnobl, benna, Jim, cfe-commits, asb
Tags: #clang
Differential Revision: https://reviews.llvm.org/D64008
llvm-svn: 364777
Summary:
The changes in D59673 made the choice redundant, since we can achieve
single-file split DWARF just by not setting an output file name.
Like llc we can also derive whether to enable Split DWARF from whether
-split-dwarf-file is set, so we don't need the flag at all anymore.
The test CodeGen/split-debug-filename.c distinguished between having set
or not set -enable-split-dwarf with -split-dwarf-file, but we can
probably just always emit the metadata into the IR.
The flag -split-dwarf wasn't used at all anymore.
Reviewers: dblaikie, echristo
Reviewed By: dblaikie
Differential Revision: https://reviews.llvm.org/D63167
llvm-svn: 364479
Summary:
This diff enables address sanitizer on Emscripten.
On Emscripten, real memory starts at the value passed to --global-base.
All memory before this is used as shadow memory, and thus the shadow mapping
function is simply dividing by 8.
Reviewers: tlively, aheejin, sbc100
Reviewed By: sbc100
Subscribers: dschuff, sbc100, jgravelle-google, hiraditya, sunfish, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D63742
llvm-svn: 364468
This patch introduces support of hip_pinned_shadow variable for HIP.
A hip_pinned_shadow variable is a global variable with attribute hip_pinned_shadow.
It has external linkage on device side and has no initializer. It has internal
linkage on host side and has initializer or static constructor. It can be accessed
in both device code and host code.
This allows HIP runtime to implement support of HIP texture reference.
Differential Revision: https://reviews.llvm.org/D62738
llvm-svn: 364381
The android target assumes that for the x86_64 target, the CPU supports SSE4.2
and popcnt. This implies that the CPU is Nehalem or newer. This should be
sufficiently new to provide the double word compare and exchange instruction.
This allows us to directly lower `__sync_val_compare_and_swap_16` to a `cmpxchg16b`.
It appears that the libatomic in android's NDK does not provide the
implementation for lowering calls to the library function.
llvm-svn: 364352
This reverts r363985 (git commit d5f16d6cfc)
This test can't use printf on Windows because the path contains
backslashes which must not be interpreted as escapes by printf.
llvm-svn: 364089
Summary:
The GCC RISC-V toolchain accepts `-msave-restore` and `-mno-save-restore`
to control whether libcalls are used for saving and restoring the stack within
prologues and epilogues.
Clang currently errors if someone passes -msave-restore or -mno-save-restore.
This means that people need to change build configurations to use clang. This
patch adds these flags, so that clang invocations can now match gcc.
As the RISC-V backend does not currently have a `save-restore` target feature,
we emit a warning if someone requests `-msave-restore`. LLVM does not error if
we pass the (unimplemented) target features `+save-restore` or `-save-restore`.
Reviewers: asb, luismarques
Reviewed By: asb
Subscribers: rbar, johnrusso, simoncook, apazos, sabuasal, niosHD, kito-cheng, shiva0217, jrtc27, zzheng, edward-jones, rogfer01, MartinMosbeck, brucehoult, the_o, rkruppe, PkmX, jocewei, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D63498
llvm-svn: 364018
Clang :: Driver/cl-response-file.c currently FAILs on Solaris:
Command Output (stderr):
--
/vol/llvm/src/clang/dist/test/Driver/cl-response-file.c:10:11: error: CHECK: expected string not found in input
// CHECK: "-I" "{{.*}}\\Inputs\\cl-response-file\\" "-D" "FOO=2"
^
Looking at the generated response file reveals that this is no surprise:
/I/vol/llvm/src/clang/dist/test/Driver\Inputs
with no newline at the end. The echo command used to create it boils down to
echo 'a\cb'
However, one cannot expect \c to be emitted literally: e.g. bash's builtin
echo has
\c suppress further output
I've tried various combinations of builtin echo, /usr/bin/echo, GNU echo if
different, the same for printf, and the backslash unescaped and quoted
(a\cb and a\\cb). The only combination that worked reliably on Solaris,
Linux, and macOS was
printf 'a\\cb'
so this is what this patch uses. Tested on amd64-pc-solaris2.11 and
x86_64-pc-linux-gnu.
Differential Revision: https://reviews.llvm.org/D63600
llvm-svn: 363985
Use -fsave-optimization-record=<format> to specify a different format
than the default, which is YAML.
For now, only YAML is supported.
llvm-svn: 363573
The flag is useful when wanting to create .o files that are independent
from the absolute path to the build directory. -fdebug-prefix-map= can
be used to the same effect, but it requires putting the absolute path
to the build directory on the build command line, so it still requires
the build command line to be dependent on the absolute path of the build
directory. With this flag, "-fdebug-compilation-dir ." makes it so that
both debug info and the compile command itself are independent of the
absolute path of the build directory, which is good for build
determinism (in the sense that the build is independent of which
directory it happens in) and for caching compile results.
(The tradeoff is that the debugger needs explicit configuration to know
the build directory. See also http://dwarfstd.org/ShowIssue.php?issue=171130.2)
Differential Revision: https://reviews.llvm.org/D63387
llvm-svn: 363548
Summary:
With Split DWARF the resulting object file (then called skeleton CU)
contains the file name of another ("DWO") file with the debug info.
This can be a problem for remote compilation, as it will contain the
name of the file on the compilation server, not on the client.
To use Split DWARF with remote compilation, one needs to either
* make sure only relative paths are used, and mirror the build directory
structure of the client on the server,
* inject the desired file name on the client directly.
Since llc already supports the latter solution, we're just copying that
over. We allow setting the actual output filename separately from the
value of the DW_AT_[GNU_]dwo_name attribute in the skeleton CU.
Fixes PR40276.
Reviewers: dblaikie, echristo, tejohnson
Reviewed By: dblaikie
Differential Revision: https://reviews.llvm.org/D59673
llvm-svn: 363496
Summary:
This is the first in a series of changes trying to align clang -cc1
flags for Split DWARF with those of llc. The unfortunate side effect of
having -split-dwarf-output for single file Split DWARF will disappear
again in a subsequent change.
The change is the result of a discussion in D59673.
Reviewers: dblaikie, echristo
Reviewed By: dblaikie
Differential Revision: https://reviews.llvm.org/D63130
llvm-svn: 363494
Add constraints for the test that require specific backend targets
to be registered.
Remove trailing whitespace in the doc.
Differential Revision: https://reviews.llvm.org/D63105
llvm-svn: 363475
This patch allows clang users to print out a list of supported CPU models using
clang [--target=<target triple>] --print-supported-cpus
Then, users can select the CPU model to compile to using
clang --target=<triple> -mcpu=<model> a.c
It is a handy feature to help cross compilation.
llvm-svn: 363464
Move include path construction from
InitHeaderSearch::AddDefaultIncludePaths in the Driver which appears
to be the more modern/correct way of doing things.
Differential Revision: https://reviews.llvm.org/D63030
llvm-svn: 363241
It breaks if build directory is not writable. This change is required
to fix our integrate.
Also add a flush() call, otherwise time trace option does not produce
the full output.
llvm-svn: 363052
Summary:
Add output to `llvm::errs()` when `-ftime-trace` option is enabled,
add regression test checking this option works as expected.
Reviewers: thakis, aganea
Subscribers: cfe-commits, llvm-commits
Tags: #clang, #llvm
This is recommit of r362821
Differential Revision: https://reviews.llvm.org/D61914
llvm-svn: 363036
Summary:
Add output to `llvm::errs()` when `-ftime-trace` option is enabled,
add regression test checking this option works as expected.
Reviewers: thakis, aganea
Subscribers: cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D61914
........
Breaks buildbots - @anton-afanasyev please can you take a look?
llvm-svn: 362816
Summary:
Add output to `llvm::errs()` when `-ftime-trace` option is enabled,
add regression test checking this option works as expected.
Reviewers: thakis, aganea
Subscribers: cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D61914
llvm-svn: 362792
Change D60691 caused some knock-on failures that weren't caught by the
existing tests. Firstly, selecting a CPU that should have had a
restricted FPU (e.g. `-mcpu=cortex-m4`, which should have 16 d-regs
and no double precision) could give the unrestricted version, because
`ARM::getFPUFeatures` returned a list of features including subtracted
ones (here `-fp64`,`-d32`), but `ARMTargetInfo::initFeatureMap` threw
away all the ones that didn't start with `+`. Secondly, the
preprocessor macros didn't reliably match the actual compilation
settings: for example, `-mfpu=softvfp` could still set `__ARM_FP` as
if hardware FP was available, because the list of features on the cc1
command line would include things like `+vfp4`,`-vfp4d16` and clang
didn't realise that one of those cancelled out the other.
I've fixed both of these issues by rewriting `ARM::getFPUFeatures` so
that it returns a list that enables every FP-related feature
compatible with the selected FPU and disables every feature not
compatible, which is more verbose but means clang doesn't have to
understand the dependency relationships between the backend features.
Meanwhile, `ARMTargetInfo::handleTargetFeatures` is testing for all
the various forms of the FP feature names, so that it won't miss cases
where it should have set `HW_FP` to feed into feature test macros.
That in turn caused an ordering problem when handling `-mcpu=foo+bar`
together with `-mfpu=something_that_turns_off_bar`. To fix that, I've
arranged that the `+bar` suffixes on the end of `-mcpu` and `-march`
cause feature names to be put into a separate vector which is
concatenated after the output of `getFPUFeatures`.
Another side effect of all this is to fix a bug where `clang -target
armv8-eabi` by itself would fail to set `__ARM_FEATURE_FMA`, even
though `armv8` (aka Arm v8-A) implies FP-Armv8 which has FMA. That was
because `HW_FP` was being set to a value including only the `FPARMV8`
bit, but that feature test macro was testing only the `VFP4FPU` bit.
Now `HW_FP` ends up with all the bits set, so it gives the right
answer.
Changes to tests included in this patch:
* `arm-target-features.c`: I had to change basically all the expected
results. (The Cortex-M4 test in there should function as a
regression test for the accidental double-precision bug.)
* `arm-mfpu.c`, `armv8.1m.main.c`: switched to using `CHECK-DAG`
everywhere so that those tests are no longer sensitive to the order
of cc1 feature options on the command line.
* `arm-acle-6.5.c`: been updated to expect the right answer to that
FMA test.
* `Preprocessor/arm-target-features.c`: added a regression test for
the `mfpu=softvfp` issue.
Reviewers: SjoerdMeijer, dmgreen, ostannard, samparker, JamesNagurne
Reviewed By: ostannard
Subscribers: srhines, javed.absar, kristof.beyls, hiraditya, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D62998
llvm-svn: 362791
Now, when clang processes an argument of the form "-march=foo+x+y+z",
then instead of calling getArchExtFeature() for each of the extension
names "x", "y", "z" and appending the returned string to its list of
low-level subtarget features, it will call appendArchExtFeatures()
which does the appending itself.
The difference is that appendArchExtFeatures can add _more_ than one
low-level feature name to the output feature list if it has to, and
also, it gets told some information about what base architecture and
CPU the extension is going to go with, which means that "+fp" can now
mean something different for different CPUs. Namely, "+fp" now selects
whatever the _default_ FPU is for the selected CPU and/or
architecture, as defined in the ARM_ARCH or ARM_CPU_NAME macros in
ARMTargetParser.def.
On the clang side, I adjust DecodeARMFeatures to call the new
appendArchExtFeatures function in place of getArchExtFeature. This
means DecodeARMFeatures needs to be passed a CPU name and an ArchKind,
which meant changing its call sites to make those available, and also
sawing getLLVMArchSuffixForARM in half so that you can get an ArchKind
enum value out of it instead of a string.
Also, I add support here for the extension name "+fp.dp", which will
automatically look through the FPU list for something that looks just
like the default FPU except for also supporting double precision.
Differential Revision: https://reviews.llvm.org/D60697
llvm-svn: 362601
Tests that use -O1, -O2 and -O3 would often produce different results
with the new pass manager which makes these tests fail. Disable new PM
explicitly for these tests.
Differential Revision: https://reviews.llvm.org/D58375
llvm-svn: 362580
Part 2 (the Clang portion) of D59881.
This patch (first of two patches) enables the vectorizer to recognize the
IBM MASS vector library routines. This patch specifically adds support for
recognizing the -vector-library=MASSV option, and defines mappings from IEEE
standard scalar math functions to generic PowerPC MASS vector counterparts.
For instance, the generic PowerPC MASS vector entry for double-precision
cbrt function is __cbrtd2_massv.
The second patch will further lower the generic PowerPC vector entries to
PowerPC subtarget-specific entries.
For instance, the PowerPC generic entry cbrtd2_massv is lowered to
cbrtd2_P9 for Power9 subtarget.
The overall support for MASS vector library is presented as such in two patches
for ease of review.
Patch by Jeeva Paudel.
Differential revision: https://reviews.llvm.org/D59881
llvm-svn: 362571
The recent change D60691 introduced a bug in clang when handling
option combinations such as `-mcpu=cortex-m4 -mfpu=none`. Those
options together should select Cortex-M4 but disable all use of
hardware FP, but in fact, now hardware FP instructions can still be
generated in that mode.
The reason is because the handling of FPUVersion::NONE disables all
the same feature names it used to, of which the base one is `vfp2`.
But now there are further features below that, like `vfp2d16fp` and
(following D60694) `fpregs`, which also need to be turned off to
disable hardware FP completely.
Added a tiny test which double-checks that compiling a simple FP
function doesn't access the FP registers.
Reviewers: SjoerdMeijer, dmgreen
Reviewed By: dmgreen
Subscribers: lebedev.ri, javed.absar, kristof.beyls, hiraditya, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D62729
llvm-svn: 362380
The new tests were failing, because I missed dependent patch D60697.
I have removed the failing cases for now, which I will restore once
D60697 is in.
llvm-svn: 362100
Given the existing infrastructure in LLVM side for +fp and +fp.dp,
this is more or less trivial, needing only one tiny source change and
a couple of tests.
Patch by Simon Tatham.
Differential Revision: https://reviews.llvm.org/D60699
llvm-svn: 362096
Modern ELF platforms use -fuse-init-array to emit .init_array instead of
.ctors . ld.bfd and gold --ctors-in-init-array merge .init_array and
.ctors into .init_array but lld doesn't do that.
If crtbegin*.o crtend*.o don't provide .ctors/.dtors, such .ctors in
user object files can lead to crash (see PR42002. The first and the last
elements in .ctors/.dtors are ignored - they are traditionally provided
by crtbegin*.o crtend*.o).
Call addClangTargetOptions() to ensure -fuse-init-array is rendered on
modern ELF platforms. On Hexagon, this renders -target-feature
+reserved-r19 for -ffixed-r19.
Reviewed By: compnerd
Differential Revision: https://reviews.llvm.org/D62509
llvm-svn: 362052
This matches gcc -static-pie. The intention is to prevent dynamic
relocations in read-only segments.
In ld.bfd and gold, -z notext is the default. If text relocations are needed:
* -z notext: allow and emit DF_TEXTREL.
DF_TEXTREL is not emitted if there is no text relocation.
* -z text: error
In lld, -z text is the default (this change is a no-op).
* -z text: error on text relocations
* -z notext: allow text relocations, and emit DF_TEXTREL no matter whether
text relocations exist.
Reviewed By: sivachandra
Differential Revision: https://reviews.llvm.org/D62606
llvm-svn: 362050
This is useful when looking for directories or files relative to the
toolchain root, e.g. include/c++/v1. This change also adds a test
to make sure this functionality doesn't regress in the future.
Differential Revision: https://reviews.llvm.org/D62558
llvm-svn: 361903
Currently the `-working-directory` option does not actually impact the working
directory for all of the clang driver, it only impacts how files are looked up
to make sure they exist. This means that that clang passes the wrong paths
to -fdebug-compilation-dir and -coverage-notes-file.
This patch fixes that by changing all the places in the driver where we convert
to absolute paths to use the VFS, and then calling setCurrentWorkingDirectory on
the VFS. This also changes the default VFS for `Driver` to use a virtualized
working directory, instead of changing the process's working directory.
Differential Revision: https://reviews.llvm.org/D62271
llvm-svn: 361885
Update the `cl` emulation to support the `/Zc:char8_t[-]?` options as per the
MSVC 2019.1 toolset. These are aliases for `-fchar8_t` and `-fno-char8_t`.
llvm-svn: 361859
Those two subtarget features were awkward because their semantics are
reversed: each one indicates the _lack_ of support for something in
the architecture, rather than the presence. As a consequence, you
don't get the behavior you want if you combine two sets of feature
bits.
Each SubtargetFeature for an FP architecture version now comes in four
versions, one for each combination of those options. So you can still
say (for example) '+vfp2' in a feature string and it will mean what
it's always meant, but there's a new string '+vfp2d16sp' meaning the
version without those extra options.
A lot of this change is just mechanically replacing positive checks
for the old features with negative checks for the new ones. But one
more interesting change is that I've rearranged getFPUFeatures() so
that the main FPU feature is appended to the output list *before*
rather than after the features derived from the Restriction field, so
that -fp64 and -d32 can override defaults added by the main feature.
Reviewers: dmgreen, samparker, SjoerdMeijer
Subscribers: srhines, javed.absar, eraman, kristof.beyls, hiraditya, zzheng, Petar.Avramovic, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D60691
llvm-svn: 361845
This is a follow up to r361432, changing the layout of per-target
runtimes to more closely resemble multiarch. While before, we used
the following layout:
[RESOURCE_DIR]/<target>/lib/libclang_rt.<runtime>.<ext>
Now we use the following layout:
[RESOURCE_DIR]/lib/<target>/libclang_rt.<runtime>.<ext>
This also more closely resembles the existing "non-per-target" layout:
[RESOURCE_DIR]/lib/<os>/libclang_rt.<runtime>-<arch>.<ext>
This change will enable further simplification of the driver logic
in follow up changes.
Differential Revision: https://reviews.llvm.org/D62469
llvm-svn: 361784
This is a follow up to r361432 and r361504 which addresses issues
introduced by those changes. Specifically, it avoids duplicating
file and runtime paths in case when the effective triple is the
same as the cannonical one. Furthermore, it fixes the broken multilib
setup in the Fuchsia driver and deduplicates some of the code.
Differential Revision: https://reviews.llvm.org/D62442
llvm-svn: 361709
New -cc1 arguments, such as -faddrsig, have started appearing after the
input name. I personally find it convenient for the input to be the last
argument to the compile command line, since I often need to edit it when
running crash reproduction scripts.
Differential Revision: https://reviews.llvm.org/D62270
llvm-svn: 361530
OptTable treats arguments starting with / that aren't a known option
as filenames. This means lld-link's and clang-cl's typo correction for
unknown flags didn't do spell checking for misspelled options that start
with /.
I first tried changing OptTable, but that got pretty messy, see PR41787
comments 2 and 3.
Instead, let lld-link's and clang's (including clang-cl's) "file not
found" diagnostic check if a non-existent file looks like it could be a
mis-spelled option, and if so add a "did you mean" suggestion to the
"file not found" diagnostic.
While here, make formatting of a few diagnostics a bit more
self-consistent.
Fixes PR41787.
Differential Revision: https://reviews.llvm.org/D62276
llvm-svn: 361518
This change is a consequence of the discussion in "RFC: Place libs in
Clang-dedicated directories", specifically the suggestion that
libunwind, libc++abi and libc++ shouldn't be using Clang resource
directory. Tools like clangd make this assumption, but this is
currently not true for the LLVM_ENABLE_PER_TARGET_RUNTIME_DIR build.
This change addresses that by moving the output of these libraries to
lib/$target/c++ and include/c++ directories, leaving resource directory
only for compiler-rt runtimes and Clang builtin headers.
Differential Revision: https://reviews.llvm.org/D59168
llvm-svn: 361432
Before e97b5f5cf3 ([clang][Darwin] Refactor header search path logic
into the driver), both --sysroot and -isysroot worked to specify where
to look for system and C++ headers on Darwin. However, that change
caused clang to start ignoring --sysroot.
This fixes the regression, and adds tests.
(I also note that on all other platforms, clang seems to almost
completely ignore -isysroot, but that's another issue...)
Differential Revision: https://reviews.llvm.org/D62268
llvm-svn: 361429
Summary:
Values returned by GCCInstallation.getParentLibPath() and
GCCInstallation.getTriple() are not valid unless
GCCInstallation.isValid() returns true. This has previously been
ignored, and the former two values were used without checking whether
GCCInstallation is valid. This led to the bad path "/../bin" being added
to the list of program paths.
author: danielmentz "Daniel Mentz <danielmentz@google.com>"
Reviewers: #clang, tstellar, srhines
Reviewed By: srhines
Subscribers: danielmentz, ormris, nickdesaulniers, srhines, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D57930
llvm-svn: 361314
Summary:
This commit moves the logic for determining system, resource and C++
header search paths from CC1 to the driver. This refactor has already
been made for several platforms, but Darwin had been left behind.
This refactor tries to implement the previous search path logic with
perfect accuracy. In particular, the order of all include paths inside
CC1 and all paths that were skipped because nonexistent are conserved
after the refactor. This change was also tested against a code base
of significant size and revealed no problems.
Reviewers: jfb, arphaman
Subscribers: nemanjai, javed.absar, kbarton, christof, jkorous, dexonsmith, jsji, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D61963
llvm-svn: 361278
Summary:
This patch modifies the AVR toolchain so that if avr-gcc and avr-libc
are detected during compilation, the CRT, libgcc, libm, and libc anre
linked.
This matches avr-gcc's default behaviour, and the expected behaviour of
all C compilers - including the C runtime.
avr-gcc also needs a -mmcu specified in order to link runtime libraries.
The difference betwen this patch and avr-gcc is that this patch will
warn users whenever they compile without a runtime, as opposed to GCC,
which silently trims the runtime libs from the linker arguments when no
-mmcu is specified.
Reviewers: aaron.ballman, kparzysz, asb, hfinkel, brucehoult, TimNN
Subscribers: cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D54334
llvm-svn: 361116
This is needed so lld-link can find clang_rt.profile when self hosting
on Windows with PGO. Using clang-cl as a linker knows to add the library
but self hosting, using -DCMAKE_LINKER=<...>/lld-link.exe doesn't.
Differential Revision: https://reviews.llvm.org/D61742
llvm-svn: 360674
This driver flag is useful when users want to link against the compiler's
builtins, but nothing else, and so use flags like -nostdlib.
Darwin can't use -nolibc & nostdlib++ like other platforms on because we
disable all runtime lib linking with -static, which we still want to have
an option to link with the builtins.
Differential Revision: https://reviews.llvm.org/D58320
llvm-svn: 360483
The test 'test/Driver/XRay/xray-instrument-os.c' is supposed to XFAIL on -darwin triples.
However, LLVM can be configured to be built with a -macos triple instead, which is equivalent
to -darwin. This commit updates the XFAIL condition to also XFAIL with a -macos host triple.
llvm-svn: 360374
Summary:
When using `clang -mcpu=CPUNAME+FEATURELIST`,
the implied features defined by CPUNAME are
not obtained, as the entire string is passed.
This fixes that by spiting the cpuname
string in the first `+`, if any.
For example, when using
```clang -### --target=arm-arm-none-eabi -march=armv7-a -mcpu=cortex-a8+nocrc```
the intrinsic
```"target-feature" "+dsp"```
implied by `cortex-a8` is missing.
Reviewers: keith.walker.arm, DavidSpickett, carwil
Reviewed By: DavidSpickett
Subscribers: javed.absar, kristof.beyls, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D61668
llvm-svn: 360324
Summary:
In this patch we propose a temporary solution to resolving math functions for the NVPTX toolchain, temporary until OpenMP variant is supported by Clang.
We intercept the inclusion of math.h and cmath headers and if we are in the OpenMP-NVPTX case, we re-use CUDA's math function resolution mechanism.
Authors:
@gtbercea
@jdoerfert
Reviewers: hfinkel, caomhin, ABataev, tra
Reviewed By: hfinkel, ABataev, tra
Subscribers: JDevlieghere, mgorny, guansong, cfe-commits, jdoerfert
Tags: #clang
Differential Revision: https://reviews.llvm.org/D61399
llvm-svn: 360265
This commit appears to be breaking stage-2 builds on GreenDragon. The
OpenMP wrappers for cmath and math.h are copied into the root of the
resource directory and cause a cyclic dependency in module 'Darwin':
Darwin -> std -> Darwin. This blows up when CMake is testing for modules
support and breaks all stage 2 module builds, including the ThinLTO bot
and all LLDB bots.
CMake Error at cmake/modules/HandleLLVMOptions.cmake:497 (message):
LLVM_ENABLE_MODULES is not supported by this compiler
llvm-svn: 360192
This matches the behavior of the old pass manager. There are some
targets that don't have target machine at all (e.g. le32, spir) which
whose tests would never run with new pass manager. Similarly, we would
need to disable tests for targets that are disabled.
Differential Revision: https://reviews.llvm.org/D58374
llvm-svn: 360100
Summary:
In this patch we propose a temporary solution to resolving math functions for the NVPTX toolchain, temporary until OpenMP variant is supported by Clang.
We intercept the inclusion of math.h and cmath headers and if we are in the OpenMP-NVPTX case, we re-use CUDA's math function resolution mechanism.
Authors:
@gtbercea
@jdoerfert
Reviewers: hfinkel, caomhin, ABataev, tra
Reviewed By: hfinkel, ABataev, tra
Subscribers: mgorny, guansong, cfe-commits, jdoerfert
Tags: #clang
Differential Revision: https://reviews.llvm.org/D61399
llvm-svn: 360063
Summary:
1. Enable infrastructure of AVX512_BF16, which is supported for BFLOAT16 in Cooper Lake;
2. Enable intrinsics for VCVTNE2PS2BF16, VCVTNEPS2BF16 and DPBF16PS instructions, which are Vector Neural Network Instructions supporting BFLOAT16 inputs and conversion instructions from IEEE single precision.
For more details about BF16 intrinsic, please refer to the latest ISE document: https://software.intel.com/en-us/download/intel-architecture-instruction-set-extensions-programming-reference
Patch by LiuTianle
Reviewers: craig.topper, smaslov, LuoYuanke, wxiao3, annita.zhang, spatel, RKSimon
Reviewed By: craig.topper
Subscribers: mgorny, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D60552
llvm-svn: 360018
When user specifies non-existent directory to -fcrash-diagnostics-dir,
create it rather than failing with an error as would be the case before.
Differential Revision: https://reviews.llvm.org/D61542
llvm-svn: 359954
Summary:
By default, `parseCommandLineOptions()` will accept either a
`-` or `--` prefix for long options -- options with names longer than
a single character.
While this change does not affect behavior, it will be helpful with a
subsequent change that requires long options use the `--` prefix.
Reviewers: rnk, thopre
Reviewed By: thopre
Subscribers: thopre, cfe-commits, hiraditya, llvm-commits
Tags: #llvm, #clang
Differential Revision: https://reviews.llvm.org/D61269
llvm-svn: 359909
This test checks whether crtbegin.o and crtend.o appear on the link
line, but names of these files may be affected by the choice of the
rtlib, specifically when compiler-rt is used as the default rtlib
the names will be clang_rt.crtbegin.o and clang_rt.crtend.o instead
of crtbeginS.o and crtendS.o. To avoid the test failure, explicitly
request to use the platform rtlib.
Differential Revision: https://reviews.llvm.org/D61383
llvm-svn: 359706
Similar to https://reviews.llvm.org/D61334, update clang tests to use the
"wasm32-wasi" triple, removing the "-musl" environment and omitting the
"-unknown" vendor.
Differential Revision: https://reviews.llvm.org/D61338
Reviewer: sbc100
llvm-svn: 359630
When compiler-rt is selected as the runtime library for Linux targets
use its crtbegin.o/crtend.o implemenetation rather than platform one
if available.
Differential Revision: https://reviews.llvm.org/D59264
llvm-svn: 359603
This introduces a support for multilibs to Fuchsia driver. Unlike the
existing multilibs that are used primarily for handling different
architecture variants, we use multilibs to handle different variants
of Clang runtime libraries: -fsanitize=address and -fno-exceptions
are the two we support initially. This replaces the existing support
for sanitized runtimes libraries that was only used by Fuchsia driver
and it also refactors some of the logic to allow sharing between GNU
and Fuchsia drivers.
Differential Revision: https://reviews.llvm.org/D61040
llvm-svn: 359360
is running in a toolchain outside of xcode
'libarclite' usually lives in the same toolchain as 'clang'. However, the
Swift open source toolchains for macOS distribute Clang without 'libarclite'.
In that case, to allow the linker to find 'libarclite', we point to the
'libarclite' that should be in the XcodeDefault toolchain instead. The
path to the toolchain is inferred from the SDK path if it's specified.
https://bugs.swift.org/browse/SR-9972
rdar://49947573
llvm-svn: 359353
The behaviour of not quoting spaces appears to have been introduced by
mistake in r190620.
Patch by Brad Moody!
Differential revision: https://reviews.llvm.org/D60997
llvm-svn: 359077
Summary:
When -gsplit-dwarf is used together with other -g options, in most cases
the computed debug info level is decided by the last -g option, with one
special case (see below). This patch drops that special case and thus
makes it easy to reason about:
// If a lower debug level -g comes after -gsplit-dwarf, in some cases
// -gsplit-dwarf is cancelled.
-gsplit-dwarf -g0 => 0
-gsplit-dwarf -gline-directives-only => DebugDirectivesOnly
-gsplit-dwarf -gmlt -fsplit-dwarf-inlining => 1
-gsplit-dwarf -gmlt -fno-split-dwarf-inlining => 1 + split
// If -gsplit-dwarf comes after -g options, with this patch, the net
// effect is 2 + split for all combinations
-g0 -gsplit-dwarf => 2 + split
-gline-directives-only -gsplit-dwarf => 2 + split
-gmlt -gsplit-dwarf -fsplit-dwarf-inlining => 2 + split
-gmlt -gsplit-dwarf -fno-split-dwarf-inlining => 1 + split (before) 2 + split (after)
The last case has been changed. In general, if the user intends to lower
debug info level, place that -g option after -gsplit-dwarf.
Some context:
In gcc, the last of -gsplit-dwarf -g0 -g1 -g2 -g3 -ggdb[0-3] -gdwarf-*
... decides the debug info level (-gsplit-dwarf -gdwarf-* have level 2).
It is a bit unfortunate that -gsplit-dwarf -gdwarf-* ... participate in
the level computation but that is the status quo.
Reviewers: dblaikie, echristo, probinson
Reviewed By: dblaikie, probinson
Subscribers: probinson, aprantl, jdoerfert, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D59923
llvm-svn: 358544
LLDB can't currently handle Clang's default (limit/no-standalone) DWARF,
so platforms that default to LLDB (Darwin) or anyone else manually
requesting LLDB tuning, should also get standalone DWARF.
That doesn't mean a user can't explicitly enable (because they have
other reasons to prefer standalone DWARF (such as that they're only
building half their application with debug info enabled, and half
without - or because they're tuning for GDB, but want to be able to use
it under LLDB too (this is the default on FreeBSD))) or disable (testing
LLDB fixes/improvements that handle no-standalone mode, building C code,
perhaps, which wouldn't have the LLDB<>no-standalone conflict, etc) the
feature regardless of the tuning.
llvm-svn: 358464
This reverts r358409, which I think broke the bots in compiler-rt.
Since I'm having trouble reproducing the failure, I'm reverting this
until I can investigate locally.
llvm-svn: 358437
Summary:
In r350649, I changed aligned allocation from being available starting
in macosx10.13 to macosx10.14. However, aligned allocation is indeed
available starting with macosx10.13, my investigation had been based
on the wrong libc++abi dylib.
This means that Clang before the fix will be more stringent when it
comes to aligned allocation -- it will not allow it when back-deploying
to macosx 10.13, when it would actually be safe to do so.
Note that a companion change will be coming to fix the libc++ tests.
Reviewers: ahatanak
Subscribers: jkorous, dexonsmith, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D60626
llvm-svn: 358409
Use -mlink-builtin-bitcode instead of llvm-link to link
device library so that device library bitcode and user
device code can be compiled in a consistent way.
This is the same approach used by CUDA and OpenMP.
Differential Revision: https://reviews.llvm.org/D60513
llvm-svn: 358290
Disabled by default as this is still an experimental feature.
Reviewed By: thakis
Differential Revision: https://reviews.llvm.org/D59221
llvm-svn: 358285
`test/Driver/debug-options.c` to ensure that the driver
selects the DWARF 2 version as intended by the test.
Fixes the `test/Driver/debug-options.c` test regression on GreenDragon
on macOS that started failing after r357713.
llvm-svn: 357740
Summary:
In the future, Android releases will support DWARF 5, but we need to
ensure that older targets only have DWARF 4 generated for them. This
patch inserts that verification for all Android releases now. The patch
also fixes 2 minor mistakes (a mistakenly moved RUN line, and the
missing G_DWARF2 check label).
Reviewers: aprantl
Reviewed By: aprantl
Subscribers: chh, pirama, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D60238
llvm-svn: 357713
When setting up library and tools paths when detecting an accompanying GCC
installation only riscv32 was handled. As a consequence when targetting
riscv64 neither the linker nor libraries would be found. This adds handling
and tests for riscv64.
Differential Revision: https://reviews.llvm.org/D53392
Patch by Edward Jones.
llvm-svn: 357699
riscv32-linux-unknown-elf was a weird thing to test for as it doesn't match
the triple used in any common RISC-V toolchain distributions (e.g.
riscv-gnu-toolchain scripts produce riscv{32,64}-unknown-linux-gnu).
llvm-svn: 357693
In case of N64 ABI toolchain paths migth have `mips-linux-gnuabi64`
or `mips-linux-gnu` directory regardless of selected environment.
Check both variants while detecting a multiarch triple.
Fix for the bug https://bugs.llvm.org/show_bug.cgi?id=41204
llvm-svn: 357506
Android does not allow shared text relocations. Enable the linker
warning to detect them by default.
Reviewers: srhines, pirama
Reviewed By: srhines
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D53344
llvm-svn: 357296
Effectively reverts r337612. The issues that cropped up with the last
attempt appear to have gone away.
Differential Revision: https://reviews.llvm.org/D59008
llvm-svn: 357285
Add an -mtp=el[0-3] option to select which of the AArch64 thread ID registers
will be used for the TLS base pointer.
This is a followup to rL356657 which added subtarget features to enable
accesses to the privileged thread ID registers.
Patch by Philip Derrin!
Differential revision: https://reviews.llvm.org/D59631
llvm-svn: 357250
In gcc, -gsplit-dwarf is handled in gcc/gcc.c as a spec
(ASM_FINAL_SPEC): objcopy --extract-dwo + objcopy --strip-dwo. In
gcc/opts.c, -gsplit_dwarf has the same semantic of a -g. Except for the
availability of the external command 'objcopy', nothing precludes the
feature working on other ELF OSes. llvm doesn't use objcopy, so it doesn't
have to exclude other OSes.
llvm-svn: 357150
Since AArch64 has default outlining behaviour, we need to make sure that
-mno-outline is actually passed along to the linker in this case. Otherwise,
it will run by default on minsize functions even when -mno-outline is specified.
Also fix the darwin-ld test for this, which wasn't actually doing anything.
llvm-svn: 357031
The RISC-V assembler needs the target ABI because it defines a flag of the ELF
file, as described in [1].
Make clang (the driver) to pass the target ABI to -cc1as in exactly the same
way it does for -cc1.
Currently -cc1as knows about -target-abi but is not handling it. Handle it and
pass it to the MC layer via MCTargetOptions.
[1] https://github.com/riscv/riscv-elf-psabi-doc/blob/master/riscv-elf.md#file-header
Differential Revision: https://reviews.llvm.org/D59298
llvm-svn: 356981
We can't (don't want to) honor the same set of "-fuse-ld" flags with
WebAssembly since the ELF linkers (ld.lld, ld.gnu, etc) don't work with
wasm object files.
Instead we implement our own linker finding logic, similar or other
non-ELF platforms like MSVC.
We've had a few issues with CLANG_DEFAULT_LINKER overriding the
WebAssembly linker which doesn't make sense since there is no generic
linker that can handle WebAssembly today.
Differential Revision: https://reviews.llvm.org/D59743
llvm-svn: 356953
Summary:
This eliminates a linker error the user might otherwise see about how
using the 'atomics' feature requires --shared-memory.
Reviewers: sbc100, aheejin
Subscribers: dschuff, jgravelle-google, sunfish, jfb, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D59721
llvm-svn: 356817
-malign-double is currently only implemented in the -cc1 interface. But its declared in Options.td so it is a driver option too. But you try to use it with the driver you'll get a message about the option being unused.
This patch teaches the driver to pass the option through to cc1 so it won't be unused. The Options.td says the option is x86 only but I didn't see any x86 specific code in its impementation in cc1 so not sure if the documentation is wrong or if I should only pass this option through the driver on x86 targets.
Differential Revision: https://reviews.llvm.org/D59624
llvm-svn: 356706
"clang++ hello.cc --rtlib=compiler-rt"
now can works without specifying additional unwind or exception
handling libraries.
This reworked version of the feature no longer modifies today's default
unwind library for compiler-rt: which is nothing. Rather, a user
can specify -DCLANG_DEFAULT_UNWINDLIB=libunwind when configuring
the compiler.
This should address the issues from the previous version.
Update tests for new --unwindlib semantics.
Differential Revision: https://reviews.llvm.org/D59109
llvm-svn: 356508
Improved some checks and moved testing of the default header
in C++ mode into the Headers folder.
Differential Revision: https://reviews.llvm.org/D59486
llvm-svn: 356450
Change the HIP Toolchain to pass the OPT_mllvm options into OPT and LLC stages. Added a lit test to verify the command args.
Reviewers: yaxunl
Differential Revision: https://reviews.llvm.org/D59316
llvm-svn: 356277
llvm-svn 356197 relanded previously failing test case max_align.c.
This commit will reland the rest of llvm-svn 356060 commit.
Differential Revision: https://reviews.llvm.org/D59048
llvm-svn: 356208
The above commit breaks the usage of PGO and LTO when -fprofile-use is
supplied without a path. This patch changes the usage of this argument
to be inline with its use in addPGOAndCoverageFlags().
Differential revision: https://reviews.llvm.org/D59304
llvm-svn: 356111
Summary:
A first pass over platform-specific properties of the C API/ABI
on AIX for both 32-bit and 64-bit modes.
This is a continuation of D18360 by Andrew Paprocki and further work by Wu Zhao.
Patch by Andus Yu
Reviewers: apaprocki, chandlerc, hubert.reinterpretcast, jasonliu,
xingxue, sfertile
Reviewed by: hubert.reinterpretcast, apaprocki, sfertile
Differential Revision: https://reviews.llvm.org/D59048
llvm-svn: 356060
Currently we have -Rpass for filtering the remarks that are displayed as
diagnostics, but when using -fsave-optimization-record, there is no way
to filter the remarks while generating them.
This adds support for filtering remarks by passes using a regex.
Ex: `clang -fsave-optimization-record -foptimization-record-passes=inline`
will only emit the remarks coming from the pass `inline`.
This adds:
* `-fsave-optimization-record` to the driver
* `-opt-record-passes` to cc1
* `-lto-pass-remarks-filter` to the LTOCodeGenerator
* `--opt-remarks-passes` to lld
* `-pass-remarks-filter` to llc, opt, llvm-lto, llvm-lto2
* `-opt-remarks-passes` to gold-plugin
Differential Revision: https://reviews.llvm.org/D59268
Original llvm-svn: 355964
llvm-svn: 355984
Currently we have -Rpass for filtering the remarks that are displayed as
diagnostics, but when using -fsave-optimization-record, there is no way
to filter the remarks while generating them.
This adds support for filtering remarks by passes using a regex.
Ex: `clang -fsave-optimization-record -foptimization-record-passes=inline`
will only emit the remarks coming from the pass `inline`.
This adds:
* `-fsave-optimization-record` to the driver
* `-opt-record-passes` to cc1
* `-lto-pass-remarks-filter` to the LTOCodeGenerator
* `--opt-remarks-passes` to lld
* `-pass-remarks-filter` to llc, opt, llvm-lto, llvm-lto2
* `-opt-remarks-passes` to gold-plugin
Differential Revision: https://reviews.llvm.org/D59268
llvm-svn: 355964
It hasn't seen active development in years, and it hasn't reached a
state where it was useful.
Remove the code until someone is interested in working on it again.
Differential Revision: https://reviews.llvm.org/D59133
llvm-svn: 355862
We will now warn about such options being unused,
which is better than the current
"no such file or directory: '/d2foo'" errors.
Note that we can still handle specific flags separately,
e.g. we were already ignoring /d2FastFail and /d2Zi+
llvm-svn: 355682
This change is a consequence of the discussion in "RFC: Place libs in
Clang-dedicated directories", specifically the suggestion that
libunwind, libc++abi and libc++ shouldn't be using Clang resource
directory. Tools like clangd make this assumption, but this is
currently not true for the LLVM_ENABLE_PER_TARGET_RUNTIME_DIR build.
This change addresses that by moving the output of these libraries to
lib/<target> and include/ directories, leaving resource directory only
for compiler-rt runtimes and Clang builtin headers.
Differential Revision: https://reviews.llvm.org/D59013
llvm-svn: 355665
When -fno-gpu-rdc is set, device code is compiled, linked, and assembled into fat binary
and embedded as string in object files. The object files are normal object files which
can be linked by host linker. In the linking stage, the object files should not be unbundled
when -fno-gpu-rdc is set since they are normal object files, not bundles. The object files
only need to be unbundled when -fgpu-rdc is set.
Currently clang always unbundles object files, disregarding -fgpu-rdc option.
This patch fixes that.
Differential Revision: https://reviews.llvm.org/D58917
llvm-svn: 355410
When -forder-file-instrumentation is on, we pass llvm flag to enable the order file instrumentation pass.
https://reviews.llvm.org/D58751
llvm-svn: 355333
Summary:
To prevent the instability of bulk-memory in the wasm backend from
blocking separate pthread testing, temporarily remove the logic that
adds -mbulk-memory in the presence of -pthread. Since browsers will
ship bulk memory before or alongside threads, this change will be
reverted as soon as bulk memory has stabilized in the backend.
Reviewers: sbc100
Subscribers: dschuff, jgravelle-google, aheejin, sunfish, jfb, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D58854
llvm-svn: 355248
Summary:
In the clang UI, replaces -mthread-model posix with -matomics as the
source of truth on threading. In the backend, replaces
-thread-model=posix with the atomics target feature, which is now
collected on the WebAssemblyTargetMachine along with all other used
features. These collected features will also be used to emit the
target features section in the future.
The default configuration for the backend is thread-model=posix and no
atomics, which was previously an invalid configuration. This change
makes the default valid because the thread model is ignored.
A side effect of this change is that objects are never emitted with
passive segments. It will instead be up to the linker to decide
whether sections should be active or passive based on whether atomics
are used in the final link.
Reviewers: aheejin, sbc100, dschuff
Subscribers: mehdi_amini, jgravelle-google, hiraditya, sunfish, steven_wu, dexonsmith, rupprecht, jfb, jdoerfert, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D58742
llvm-svn: 355112
Remove comments and tests about passing -mcode-object-v3 to driver since it does
not work. Other -m options are OK.
Also put back -mattr=-code-object-v3 since HIP is still not ready for code object
v3.
Differential Revision: https://reviews.llvm.org/D57977
llvm-svn: 355106
This patch enables the following
1) AMD family 17h "znver2" tune flag (-march, -mcpu).
2) ISAs that are enabled for "znver2" architecture.
3) For the time being, it uses the znver1 scheduler model.
4) Tests are updated.
5) This patch is the clang counterpart to D58343
Reviewers: craig.topper
Tags: #clang
Differential Revision: https://reviews.llvm.org/D58344
llvm-svn: 354899
Summary: This change mimics GCC's support for the "-static-pie" argument.
Subscribers: cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D58307
llvm-svn: 354502
A faster way to reduce the values in teams reductions was found, the
codegen is updated to use this faster algorithm and new runtime functions.
llvm-svn: 354479
This can be used to disable libc linking. This flag is supported by
GCC since version 9 as well as some Clang target toolchains.
Differential Revision: https://reviews.llvm.org/D58326
llvm-svn: 354210
This can be used to disable libc linking. This flag is supported by
GCC since version 9 as well as some Clang target toolchains. This
change also includes tests for all -no* flags which previously weren't
covered.
Differential Revision: https://reviews.llvm.org/D58325
llvm-svn: 354208
Summary:
There are an insignificant number of ARM Android devices that don't
support NEON. Default to using NEON since that will improve
performance on the majority of devices. Users that need to target
non-NEON devices can still explicitly disable NEON.
Reviewers: srhines, pirama, kristof.beyls
Reviewed By: pirama
Subscribers: efriedma, javed.absar, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D58153
llvm-svn: 354166
Instead of letting a program fail at runtime, emit an error during
compilation.
rdar://problem/12206955
Reviewers: dexonsmith, bob.wilson, steven_wu
Reviewed By: steven_wu
Subscribers: jkorous, cfe-commits
Differential Revision: https://reviews.llvm.org/D57991
llvm-svn: 354084
The CHECK lines as structured were requiring them to appear only in a certain
position while all that is really needed is to check that they are present.
llvm-svn: 354001
This is a follow up to D48580 and D48581 which allows reserving
arbitrary general purpose registers with the exception of registers
with special purpose (X8, X16-X18, X29, X30) and registers used by LLVM
(X0, X19). This change also generalizes some of the existing logic to
rely entirely on values generated from tablegen.
Differential Revision: https://reviews.llvm.org/D56305
llvm-svn: 353957
Allow the compile options for -m such as -mxnack/-mno-xnack, -msram-ecc/-mno-sram-ecc, -mcode-object-v3/-mno-code-object-v3 to propagate into LLC args. Fix an issue where -mattr was pushed even when it was empty.
Also add lit tests to verify features are properly passed.
Differential Revision: https://reviews.llvm.org/D57977
Reviewers: yaxunl, kzhuravl
llvm-svn: 353952
Force -fuse-ld=ld, as some other tests in the same file do.
Loosen the regex matching the linker tool name as well, as this
can end up being <triple>-ld in case such a named tool exists.
llvm-svn: 353946
Profiling still doesn't seem to work properly, but this at least
hooks up the library and eases completing whatever is missing.
Differential Revision: https://reviews.llvm.org/D58107
llvm-svn: 353917
Since we removed changed the way HIP Toolchain will propagate -m options into LLC, we need to remove from these older tests.
This is related to rC353880.
Differential Revision: https://reviews.llvm.org/D57977
llvm-svn: 353885
Allow the compile options for -m such as -mxnack/-mno-xnack, -msram-ecc/-mno-sram-ecc, -mcode-object-v3/-mno-code-object-v3 to propagate into LLC args.
Also add lit tests to verify features are properly passed.
Differential Revision: https://reviews.llvm.org/D57977
Reviewers: yaxunl, kzhuravl
llvm-svn: 353880
Summary:
There have been three options related to threads and users had to set
all three of them separately to get the correct compilation results.
This makes sure the relationship between the options makes sense and
sets necessary options for users if only part of the necessary options
are specified. This does:
- Remove `-matomics`; this option alone does not enable anything, so
removed it to not confuse users.
- `-mthread-model posix` sets `-target-feature +atomics`
- `-pthread` sets both `-target-feature +atomics` and
`-mthread-model posix`
Also errors out when explicitly given options don't match, such as
`-pthread` is given with `-mthread-model single`.
Reviewers: dschuff, sbc100, tlively, sunfish
Subscribers: jgravelle-google, jfb, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D57874
llvm-svn: 353761
Instead of calling CUDA runtime to arrange function arguments,
the new API constructs arguments in a local array and the kernels
are launched with __cudaLaunchKernel().
The old API has been deprecated and is expected to go away
in the next CUDA release.
Differential Revision: https://reviews.llvm.org/D57488
llvm-svn: 352799
..and use it to control that parts of CUDA compilation
that depend on the specific version of CUDA SDK.
This patch has a placeholder for a 'new launch API' support
which is in a separate patch. The list will be further
extended in the upcoming patch to support CUDA-10.1.
Differential Revision: https://reviews.llvm.org/D57487
llvm-svn: 352798
Introduce an option to request global visibility settings be applied to
declarations without a definition or an explicit visibility, rather than
the existing behavior of giving these default visibility. When the
visibility of all or most extern definitions are known this allows for
the same optimisations -fvisibility permits without updating source code
to annotate all declarations.
Differential Revision: https://reviews.llvm.org/D56868
llvm-svn: 352391
Add tests that arguments for enabling/disabling
sb and predres are correctly being or not passed
by the driver.
Differential Revision: https://reviews.llvm.org/D57060
llvm-svn: 352203
The /AI flag is for #using directives, which I don't think we support.
This is consistent with how the /I flag is handled by MSVC. Add a test
for it.
Differential Revision: https://reviews.llvm.org/D57189
llvm-svn: 352119
These two options enable/disable emission of R_{MICRO}MIPS_JALR fixups along
with PIC calls. The linker may then try to turn PIC calls into direct jumps.
By default, these fixups do get emitted by the backend, use
'-mno-relax-pic-calls' to omit them.
Differential revision: https://reviews.llvm.org/D56878
llvm-svn: 351579
Implements PR40180.
clang-cl has one minor behavior difference with cl with this change.
Clang allows the user to enable the C++17 feature of aligned allocation
without enabling all of C++17, but MSVC will not call the aligned
allocation overloads unless -std:c++17 is passed. While our behavior is
technically incompatible, it would require making driver mode specific
changes to match MSVC precisely, and clang's behavior is useful because
it allows people to experiment with new C++17 features individually.
Therefore, I plan to leave it as is.
llvm-svn: 351249
This is an initial implementation for msp430 toolchain including
-mmcu option support
-mhwmult options support
-integrated-as by default
The toolchain uses msp430-elf-as as a linker and supports msp430-gcc toolchain tree.
Patch by Kristina Bessonova!
Differential Revision: https://reviews.llvm.org/D56658
llvm-svn: 351228
This adds support for multilib paths for wasm32 targets, following
[Debian's Multiarch conventions], and also adds an experimental OS name in
order to test it.
[Debian's Multiarch conventions]: https://wiki.debian.org/Multiarch/
Differential Revision: https://reviews.llvm.org/D56553
llvm-svn: 351164
Summary:
After r327851, Driver::GetTemporaryPath will create the file rather than
just create a potientially unqine filename. If clang driver pass the
file as parameter as -object_path_lto, ld64 will pass it back to libLTO
as GeneratedObjectsDirectory, which is going to cause a LLVM ERROR if it
is not a directory.
Now during thinLTO, pass a temp directory path to linker instread.
rdar://problem/47194182
Reviewers: arphaman, dexonsmith
Reviewed By: arphaman
Subscribers: mehdi_amini, inglorion, jkorous, cfe-commits
Differential Revision: https://reviews.llvm.org/D56608
llvm-svn: 350970
Summary:
Adds a new -f[no]split-lto-unit flag that is disabled by default to
control module splitting during ThinLTO. It is automatically enabled
for -fsanitize=cfi and -fwhole-program-vtables.
The new EnableSplitLTOUnit codegen flag is passed down to llvm
via a new module flag of the same name.
Depends on D53890.
Reviewers: pcc
Subscribers: ormris, mehdi_amini, inglorion, eraman, steven_wu, dexonsmith, cfe-commits, llvm-commits
Differential Revision: https://reviews.llvm.org/D53891
llvm-svn: 350949
Summary:
By using '..' instead of fs::parent_path.
The intention of the code was to go from 'path/to/clang/bin' to
'path/to/clang/include'. In most cases parent_path works, however it
would fail when clang is run as './clang'.
This was noticed in Chromium's bug tracker, see
https://bugs.chromium.org/p/chromium/issues/detail?id=919761
Reviewers: arphaman, thakis, EricWF
Reviewed By: arphaman, thakis
Subscribers: christof, cfe-commits
Differential Revision: https://reviews.llvm.org/D56446
llvm-svn: 350714
Summary:
r306722 added diagnostics when aligned allocation is used with deployment
targets that do not support it, but the first macosx supporting aligned
allocation was incorrectly set to 10.13. In reality, the dylib shipped
with macosx10.13 does not support aligned allocation, but the dylib
shipped with macosx10.14 does.
Reviewers: ahatanak
Subscribers: christof, jkorous, dexonsmith, cfe-commits
Differential Revision: https://reviews.llvm.org/D56445
llvm-svn: 350649
The problem is similar to D55986 but for threads: a process with the
interceptor hwasan library loaded might have some threads started by
instrumented libraries and some by uninstrumented libraries, and we
need to be able to run instrumented code on the latter.
The solution is to perform per-thread initialization lazily. If a
function needs to access shadow memory or add itself to the per-thread
ring buffer its prologue checks to see whether the value in the
sanitizer TLS slot is null, and if so it calls __hwasan_thread_enter
and reloads from the TLS slot. The runtime does the same thing if it
needs to access this data structure.
This change means that the code generator needs to know whether we
are targeting the interceptor runtime, since we don't want to pay
the cost of lazy initialization when targeting a platform with native
hwasan support. A flag -fsanitize-hwaddress-abi={interceptor,platform}
has been introduced for selecting the runtime ABI to target. The
default ABI is set to interceptor since it's assumed that it will
be more common that users will be compiling application code than
platform code.
Because we can no longer assume that the TLS slot is initialized,
the pthread_create interceptor is no longer necessary, so it has
been removed.
Ideally, lazy initialization should only cost one instruction in the
hot path, but at present the call may cause us to spill arguments
to the stack, which means more instructions in the hot path (or
theoretically in the cold path if the spills are moved with shrink
wrapping). With an appropriately chosen calling convention for
the per-thread initialization function (TODO) the hot path should
always need just one instruction and the cold path should need two
instructions with no spilling required.
Differential Revision: https://reviews.llvm.org/D56038
llvm-svn: 350429
The offload bundler action should not unbundle the input file types that does not match the action type. This fixes an issue where .so files are unbundled when the action type is object files.
Reviewers: yaxunl
Differential Revision: https://reviews.llvm.org/D56321
llvm-svn: 350426
The offload bundler action should not unbundle the input file types that does not match the action type. This fixes an issue where .so files are unbundled when the action type is object files.
llvm-svn: 350425
For some reason, the cmake build on my macbook has
LLVM_HOST_TRIPLE:STRING=i386-apple-darwin16.7.0 .
test/Driver/darwin-sdk-version.c assumed that the host triple is 64-bit, so
make it resilient against 32-bit host triples.
Differential Revision: https://reviews.llvm.org/D56067
llvm-svn: 350278
midl invokes the compiler on .idl files with /E. Before this change, we
would treat unrecognized inputs as object files. Now we pre-process to
stdout as expected. I checked that MSVC defines __cplusplus when invoked
this way, so treating the input as C++ seems like the right thing to do.
After this change, I was able to run midl like this with clang-cl:
$ midl -cpp_cmd clang-cl.exe foo.idl
Things worked for the example IDL file in the Microsoft documentation,
but beyond that, I don't know if this will work well.
Fixes PR40140
llvm-svn: 350072
Gentoo supports combining clang toolchain with GNU binutils, and many
users actually do that. As -faddrsig is not supported by GNU strip,
this results in a lot of warnings. Disable it by default and let users
enable it explicitly if they want it; with the intent of reevaluating
when the underlying feature becomes standarized.
See also: https://bugs.gentoo.org/667854
Differential Revision: https://reviews.llvm.org/D56047
llvm-svn: 350028
NFC for targets other than PS4.
Respect -nostdlib and -nodefaultlibs when enabling asan or ubsan.
Differential Revision: https://reviews.llvm.org/D55712
llvm-svn: 349508
For targets where SEH exceptions are used by default (on MinGW,
only x86_64 so far), -munwind-tables are added automatically. If
-fseh-exeptions is enabled on a target where SEH exeptions are
availble but not enabled by default yet (aarch64), we need to
pass -munwind-tables if -fseh-exceptions was specified.
Differential Revision: https://reviews.llvm.org/D55749
llvm-svn: 349452
Summary:
Add an option to initialize automatic variables with either a pattern or with
zeroes. The default is still that automatic variables are uninitialized. Also
add attributes to request uninitialized on a per-variable basis, mainly to disable
initialization of large stack arrays when deemed too expensive.
This isn't meant to change the semantics of C and C++. Rather, it's meant to be
a last-resort when programmers inadvertently have some undefined behavior in
their code. This patch aims to make undefined behavior hurt less, which
security-minded people will be very happy about. Notably, this means that
there's no inadvertent information leak when:
- The compiler re-uses stack slots, and a value is used uninitialized.
- The compiler re-uses a register, and a value is used uninitialized.
- Stack structs / arrays / unions with padding are copied.
This patch only addresses stack and register information leaks. There's many
more infoleaks that we could address, and much more undefined behavior that
could be tamed. Let's keep this patch focused, and I'm happy to address related
issues elsewhere.
To keep the patch simple, only some `undef` is removed for now, see
`replaceUndef`. The padding-related infoleaks are therefore not all gone yet.
This will be addressed in a follow-up, mainly because addressing padding-related
leaks should be a stand-alone option which is implied by variable
initialization.
There are three options when it comes to automatic variable initialization:
0. Uninitialized
This is C and C++'s default. It's not changing. Depending on code
generation, a programmer who runs into undefined behavior by using an
uninialized automatic variable may observe any previous value (including
program secrets), or any value which the compiler saw fit to materialize on
the stack or in a register (this could be to synthesize an immediate, to
refer to code or data locations, to generate cookies, etc).
1. Pattern initialization
This is the recommended initialization approach. Pattern initialization's
goal is to initialize automatic variables with values which will likely
transform logic bugs into crashes down the line, are easily recognizable in
a crash dump, without being values which programmers can rely on for useful
program semantics. At the same time, pattern initialization tries to
generate code which will optimize well. You'll find the following details in
`patternFor`:
- Integers are initialized with repeated 0xAA bytes (infinite scream).
- Vectors of integers are also initialized with infinite scream.
- Pointers are initialized with infinite scream on 64-bit platforms because
it's an unmappable pointer value on architectures I'm aware of. Pointers
are initialize to 0x000000AA (small scream) on 32-bit platforms because
32-bit platforms don't consistently offer unmappable pages. When they do
it's usually the zero page. As people try this out, I expect that we'll
want to allow different platforms to customize this, let's do so later.
- Vectors of pointers are initialized the same way pointers are.
- Floating point values and vectors are initialized with a negative quiet
NaN with repeated 0xFF payload (e.g. 0xffffffff and 0xffffffffffffffff).
NaNs are nice (here, anways) because they propagate on arithmetic, making
it more likely that entire computations become NaN when a single
uninitialized value sneaks in.
- Arrays are initialized to their homogeneous elements' initialization
value, repeated. Stack-based Variable-Length Arrays (VLAs) are
runtime-initialized to the allocated size (no effort is made for negative
size, but zero-sized VLAs are untouched even if technically undefined).
- Structs are initialized to their heterogeneous element's initialization
values. Zero-size structs are initialized as 0xAA since they're allocated
a single byte.
- Unions are initialized using the initialization for the largest member of
the union.
Expect the values used for pattern initialization to change over time, as we
refine heuristics (both for performance and security). The goal is truly to
avoid injecting semantics into undefined behavior, and we should be
comfortable changing these values when there's a worthwhile point in doing
so.
Why so much infinite scream? Repeated byte patterns tend to be easy to
synthesize on most architectures, and otherwise memset is usually very
efficient. For values which aren't entirely repeated byte patterns, LLVM
will often generate code which does memset + a few stores.
2. Zero initialization
Zero initialize all values. This has the unfortunate side-effect of
providing semantics to otherwise undefined behavior, programs therefore
might start to rely on this behavior, and that's sad. However, some
programmers believe that pattern initialization is too expensive for them,
and data might show that they're right. The only way to make these
programmers wrong is to offer zero-initialization as an option, figure out
where they are right, and optimize the compiler into submission. Until the
compiler provides acceptable performance for all security-minded code, zero
initialization is a useful (if blunt) tool.
I've been asked for a fourth initialization option: user-provided byte value.
This might be useful, and can easily be added later.
Why is an out-of band initialization mecanism desired? We could instead use
-Wuninitialized! Indeed we could, but then we're forcing the programmer to
provide semantics for something which doesn't actually have any (it's
uninitialized!). It's then unclear whether `int derp = 0;` lends meaning to `0`,
or whether it's just there to shut that warning up. It's also way easier to use
a compiler flag than it is to manually and intelligently initialize all values
in a program.
Why not just rely on static analysis? Because it cannot reason about all dynamic
code paths effectively, and it has false positives. It's a great tool, could get
even better, but it's simply incapable of catching all uses of uninitialized
values.
Why not just rely on memory sanitizer? Because it's not universally available,
has a 3x performance cost, and shouldn't be deployed in production. Again, it's
a great tool, it'll find the dynamic uses of uninitialized variables that your
test coverage hits, but it won't find the ones that you encounter in production.
What's the performance like? Not too bad! Previous publications [0] have cited
2.7 to 4.5% averages. We've commmitted a few patches over the last few months to
address specific regressions, both in code size and performance. In all cases,
the optimizations are generally useful, but variable initialization benefits
from them a lot more than regular code does. We've got a handful of other
optimizations in mind, but the code is in good enough shape and has found enough
latent issues that it's a good time to get the change reviewed, checked in, and
have others kick the tires. We'll continue reducing overheads as we try this out
on diverse codebases.
Is it a good idea? Security-minded folks think so, and apparently so does the
Microsoft Visual Studio team [1] who say "Between 2017 and mid 2018, this
feature would have killed 49 MSRC cases that involved uninitialized struct data
leaking across a trust boundary. It would have also mitigated a number of bugs
involving uninitialized struct data being used directly.". They seem to use pure
zero initialization, and claim to have taken the overheads down to within noise.
Don't just trust Microsoft though, here's another relevant person asking for
this [2]. It's been proposed for GCC [3] and LLVM [4] before.
What are the caveats? A few!
- Variables declared in unreachable code, and used later, aren't initialized.
This goto, Duff's device, other objectionable uses of switch. This should
instead be a hard-error in any serious codebase.
- Volatile stack variables are still weird. That's pre-existing, it's really
the language's fault and this patch keeps it weird. We should deprecate
volatile [5].
- As noted above, padding isn't fully handled yet.
I don't think these caveats make the patch untenable because they can be
addressed separately.
Should this be on by default? Maybe, in some circumstances. It's a conversation
we can have when we've tried it out sufficiently, and we're confident that we've
eliminated enough of the overheads that most codebases would want to opt-in.
Let's keep our precious undefined behavior until that point in time.
How do I use it:
1. On the command-line:
-ftrivial-auto-var-init=uninitialized (the default)
-ftrivial-auto-var-init=pattern
-ftrivial-auto-var-init=zero -enable-trivial-auto-var-init-zero-knowing-it-will-be-removed-from-clang
2. Using an attribute:
int dont_initialize_me __attribute((uninitialized));
[0]: https://users.elis.ugent.be/~jsartor/researchDocs/OOPSLA2011Zero-submit.pdf
[1]: https://twitter.com/JosephBialek/status/1062774315098112001
[2]: https://outflux.net/slides/2018/lss/danger.pdf
[3]: https://gcc.gnu.org/ml/gcc-patches/2014-06/msg00615.html
[4]: 776a0955ef
[5]: http://wg21.link/p1152
I've also posted an RFC to cfe-dev: http://lists.llvm.org/pipermail/cfe-dev/2018-November/060172.html
<rdar://problem/39131435>
Reviewers: pcc, kcc, rsmith
Subscribers: JDevlieghere, jkorous, dexonsmith, cfe-commits
Differential Revision: https://reviews.llvm.org/D54604
llvm-svn: 349442
The test test/Driver/darwin-sdk-version.c from r349380 checks if the macOS
deployment target can be correctly inferred from the SDK version. When the
SDK version is > host version, the driver will pick the host version, so
the old test failed on macOS < 10.14. This commit makes this test more
resilient by using an older SDK version.
llvm-svn: 349393
is not specified
The -target option allows the user to specify the build target using LLVM
triple. The triple includes the arch, and so the -arch option is redundant.
This should work just as well without the -arch. However, the driver has a bug
in which it doesn't target the "Cyclone" CPU for darwin if -target is used
without -arch. This commit fixes this issue.
rdar://46743182
Differential Revision: https://reviews.llvm.org/D55731
llvm-svn: 349382
On Darwin, using '-arch x86_64h' would always override the option passed
through '-march'.
This patch allows users to use '-march' with x86_64h, while keeping the
default to 'core-avx2'
Differential Revision: https://reviews.llvm.org/D55775
llvm-svn: 349381
pass in the -target-sdk-version to the compiler and backend
This commit adds support for reading the SDKSettings.json file in the Darwin
driver. This file is used by the driver to determine the SDK's version, and it
uses that information to pass it down to the compiler using the new
-target-sdk-version= option. This option is then used to set the appropriate
SDK Version module metadata introduced in r349119.
Note: I had to adjust the two ast tests as the SDKROOT environment variable
on macOS caused SDK version to be picked up for the compilation of source file
but not the AST.
rdar://45774000
Differential Revision: https://reviews.llvm.org/D55673
llvm-svn: 349380
Implement options in clang to enable recording the driver command-line
in an ELF section.
Implement a new special named metadata, llvm.commandline, to support
frontends embedding their command-line options in IR/ASM/ELF.
This differs from the GCC implementation in some key ways:
* In GCC there is only one command-line possible per compilation-unit,
in LLVM it mirrors llvm.ident and multiple are allowed.
* In GCC individual options are separated by NULL bytes, in LLVM entire
command-lines are separated by NULL bytes. The advantage of the GCC
approach is to clearly delineate options in the face of embedded
spaces. The advantage of the LLVM approach is to support merging
multiple command-lines unambiguously, while handling embedded spaces
with escaping.
Differential Revision: https://reviews.llvm.org/D54487
Clang Differential Revision: https://reviews.llvm.org/D54489
llvm-svn: 349155
Alexey Bataev