__fixunstfti converts a long double (IBM double-double) to an unsigned 128 bit
integer. This patch enables it to handle a previously unhandled case in which
a negative low double may impact the result of the conversion.
Collaborated with @masoud.ataei and @renenkel.
Patch By: Baptiste Saleil
Differential Revision: https://reviews.llvm.org/D69193
Summary:
Previously the CMake code looked for filepaths of the form
`<arch>/<filename>` as an indication that `<arch>/<filename>` provided a
specialization of a top-level file `<filename>`. For powerpc there was a
bug because the powerpc specialized implementations lived in `ppc/` but
the architectures were `powerpc64` and `powerpc64le` which meant that
CMake was looking for files at `powerpc64/<filename>` and
`powerpc64le/<filename>`.
The result of this is that for powerpc the builtins library contained a
duplicate symbol for `divtc3` because it had the generic implementation
and the specialized version in the built static library.
Although we could just add similar code to what there is for arm (i.e.
compute `${_arch}`) to fix this, this is extremely error prone (until
r375150 no error was raised). Instead this patch takes a different
approach that removes looking for the architecture name entirely.
Instead this patch uses the convention that a source file in a
sub-directory might be a specialization of a generic implementation and
if a source file of the same name (ignoring extension) exists at the
top-level then it is the corresponding generic implementation. This
approach is much simpler because it doesn't require keeping track of
different architecture names.
This convention already existed in repository but previously it was
implicit. This change makes it explicit.
This patch is motivated by wanting to revert r375162 which worked around
the powerpc bug found when r375150 landed.
Once it lands we should revert r375162.
Reviewers: phosek, beanz, compnerd, shiva0217, amyk, rupprecht, kongyi, mstorsjo, t.p.northover, weimingz, jroelofs, joerg, sidneym
Subscribers: nemanjai, mgorny, kristof.beyls, jsji, shchenz, steven.zhang, #sanitizers, llvm-commits
Tags: #llvm, #sanitizers
Differential Revision: https://reviews.llvm.org/D69189
Use new control bits CTR_EL0.DIC and CTR_EL0.IDC to discover the d-cache
cleaning and i-cache invalidation requirements for instruction-to-data
coherence. This matches the behavior in the latest libgcc.
Author: Shaokun Zhang <zhangshaokun@hisilicon.com>
Reviewed By: peter.smith
Differential Revision: https://reviews.llvm.org/D69247
Differential Revision: https://reviews.llvm.org/D67937
M lib/builtins/fp_add_impl.inc
M lib/builtins/fp_lib.h
M lib/builtins/fp_trunc_impl.inc
llvm-svn: 372684
Most builtins accepting or returning long double use the fp_t typedef.
Change the remaining few cases to do so.
Differential Revision: https://reviews.llvm.org/D35034
llvm-svn: 371400
compiler-rt's builtin library has generic implementations of many
functions, and then per-arch optimized implementations of some.
In the CMake build, both filter_builtin_sources() and an explicit loop
at the end of the build file (see D37166) filter out the generic
versions if a per-arch file is present.
The GN build wasn't doing this filtering. Just do the filtering manually
and explicitly, instead of being clever.
While here, also remove files from the mingw/arm build that are
redundantly listed after D39938 / r318139 (both from the CMake and the
GN build).
While here, also fix a target_os -> target_cpu typo.
Differential Revision: https://reviews.llvm.org/D65512
llvm-svn: 367448
While working on https://reviews.llvm.org/D40900, two tests were failing since __clear_cache
aborted. While libgcc's __clear_cache is just empty, this only happens because
gcc (in gcc/config/sparc/sparc.c (sparc32_initialize_trampoline, sparc64_initialize_trampoline))
emits flush insns directly.
The following patch mimics that.
Tested on sparcv9-sun-solaris2.11.
Differential Revision: https://reviews.llvm.org/D64496
llvm-svn: 366822
This assembly is part of a macro that was reformatted in D60351.
The missing space between push and { results in:
Error: bad instruction `push{r4, r5,r6,lr}'
llvm-svn: 365957
This patch enables compiler-rt on SPARC targets. Most of the changes are straightforward:
- Add 32 and 64-bit sparc to compiler-rt
- lib/builtins/fp_lib.h needed to check if the int128_t and uint128_t types exist (which they don't on sparc)
There's one issue of note: many asan tests fail to compile on Solaris/SPARC:
fatal error: error in backend: Function "_ZN7testing8internal16BoolFromGTestEnvEPKcb": over-aligned dynamic alloca not supported.
Therefore, while asan is still built, both asan and ubsan-with-asan testing is disabled. The
goal is to check if asan keeps compiling on Solaris/SPARC. This serves asan in gcc,
which doesn't have the problem above and works just fine.
With this patch, sparcv9-sun-solaris2.11 test results are pretty good:
Failing Tests (9):
Builtins-sparc-sunos :: divtc3_test.c
Builtins-sparcv9-sunos :: compiler_rt_logbl_test.c
Builtins-sparcv9-sunos :: divtc3_test.c
[...]
UBSan-Standalone-sparc :: TestCases/TypeCheck/misaligned.cpp
UBSan-Standalone-sparcv9 :: TestCases/TypeCheck/misaligned.cpp
The builtin failures are due to Bugs 42493 and 42496. The tree contained a few additonal
patches either currently in review or about to be submitted.
Tested on sparcv9-sun-solaris2.11.
Differential Revision: https://reviews.llvm.org/D40943
llvm-svn: 365880
compiler-rt already uses libtool instead of ar when building for
Apple platform, but that's not being used when builtins are being
built separately e.g. as part of the runtimes build. This change
extracts the logic setting up libtool into a separate file and uses
it from both the compiler-rt and standalone builtins build.
Differential Revision: https://reviews.llvm.org/D62820
llvm-svn: 362466
If `COMPILER_RT_ARMHF_TARGET` is set , the definition of the AEABI runtime
function `__aeabi_fcmpun` is misspelt: `__aeabi_fcmpum` instead of
`__aeabi_fcmpun`.
Patch by Konstantin Schwarz!
llvm-svn: 362424
This should hopefully address the error we're seeing in older versions
of Clang.
Differential Revision: https://reviews.llvm.org/D62554
llvm-svn: 361909
The only difference between __eqsf2 and __gtsf2 is whether they return
1 or -1 on NaN. Rather than duplicating all the code, use a macro to
define the function twice and use an argument to decide whether to
negate the return value.
Differential Revision: https://reviews.llvm.org/D61919
llvm-svn: 361207
Symbol aliases are supported by all platforms that compiler-rt builtins
target, and we can use these instead of function redirects to avoid the
extra indirection.
This is part of the cleanup proposed in "[RFC] compiler-rt builtins
cleanup and refactoring".
Differential Revision: https://reviews.llvm.org/D60931
llvm-svn: 359413
Use the uniform single line C++/99 style for code comments.
This is part of the cleanup proposed in "[RFC] compiler-rt builtins
cleanup and refactoring".
Differential Revision: https://reviews.llvm.org/D60352
llvm-svn: 359411
Update formatting to use the LLVM style.
This is part of the cleanup proposed in "[RFC] compiler-rt builtins
cleanup and refactoring".
Differential Revision: https://reviews.llvm.org/D60351
llvm-svn: 359410
The generic version of comparesf2 defines __cmpsf2 alias for libgcc
compatibility, but the ARM overlay is missing the alias.
Differential Revision: https://reviews.llvm.org/D60805
llvm-svn: 358542
This is support for building compiler-rt builtins, The library build
should be complete for a subset of supported platforms, but not all
CMake options have been replicated in GN.
We always use the just built compiler to build all the runtimes, which
is equivalent to the CMake runtimes build. This simplifies the build
configuration because we don't need to support arbitrary host compiler
and can always assume the latest Clang. With GN's toolchain support,
this is significantly more efficient than the CMake runtimes build.
Differential Revision: https://reviews.llvm.org/D60331
llvm-svn: 357821
This avoids a conflict between stdbool.h, which defines bool to _Bool in
xkeycheck.h. From what I can tell, ymath.h is an internal header, and
the intention is that users should include math.h directly instead. It
doesn't appear to provide declarations of anything required for our
builtins. This include was added back in r249513 from 2015, and it's
possible that ymath.h provided something this code needed at the time,
but today it does not.
llvm-svn: 357728
Implement rounding mode support for addxf3/subxf3.
On architectures that implemented the support, this will access the
corresponding floating point environment register to apply the
correct rounding. For other architectures, it will keep the current
behaviour and use IEEE-754 default rounding mode (to nearest, ties
to even).
ARM32/AArch64 support implemented in this change. i386 and AMD64
will be added in a follow up change.
Differential Revision: https://reviews.llvm.org/D57143
llvm-svn: 357035
Also hide __cpu_inicator_init and __cpu_features2
for similar reasons.
Summary: Make __cpu_model a hidden symbol, to match libgcc.
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D59561
llvm-svn: 356581
We were treating certain edge cases that are actually normal as denormal
results, and flushing them to zero; we shouldn't do that. Not sure this
is the cleanest way to implement this edge case, but I wanted to avoid
adding any code on the common path.
Differential Revision: https://reviews.llvm.org/D59070
llvm-svn: 356529
For 'cascadelake' this is adding a 'avx512vnni' feature check to the 0x55 skylake-avx512 model check. These CPUs use the same model number and only differ in the stepping number. But the feature flag is simpler than collecting all the stepping numbers.
For 'znver2' this is just syncing with LLVM's Host.cpp.
llvm-svn: 354927
Bionic libc relies on an old libgcc behaviour which does not set hidden
visibility attribute. Keep exporting these symbols on Android for
compatibility.
Differential Revision: https://reviews.llvm.org/D56977
llvm-svn: 351915
to reflect the new license. These used slightly different spellings that
defeated my regular expressions.
We understand that people may be surprised that we're moving the header
entirely to discuss the new license. We checked this carefully with the
Foundation's lawyer and we believe this is the correct approach.
Essentially, all code in the project is now made available by the LLVM
project under our new license, so you will see that the license headers
include that license only. Some of our contributors have contributed
code under our old license, and accordingly, we have retained a copy of
our old license notice in the top-level files in each project and
repository.
llvm-svn: 351648
to reflect the new license.
We understand that people may be surprised that we're moving the header
entirely to discuss the new license. We checked this carefully with the
Foundation's lawyer and we believe this is the correct approach.
Essentially, all code in the project is now made available by the LLVM
project under our new license, so you will see that the license headers
include that license only. Some of our contributors have contributed
code under our old license, and accordingly, we have retained a copy of
our old license notice in the top-level files in each project and
repository.
llvm-svn: 351636
This patch implements the long double __floattitf (int128_t) method for
PowerPC -- specifically to convert a 128 bit integer into a long double
(IBM double-double).
To invoke this method, one can do so by linking against compiler-rt, via the
--rtlib=compiler-rt command line option supplied to clang.
Differential Revision: https://reviews.llvm.org/D54313/
llvm-svn: 350818
This patch implements the __uint128_t __fixunstfti (long double) method for
PowerPC -- specifically to convert a long double (IBM double-double) to an
unsigned 128 bit integer.
The general approach of this algorithm is to convert the high and low doubles
of the long double and add them together if the doubles fit within 64 bits.
However, additional adjustments and scaling is performed when the high or low
double does not fit within a 64 bit integer.
To invoke this method, one can do so by linking against compiler-rt, via the
--rtlib=compiler-rt command line option supplied to clang.
Differential Revision: https://reviews.llvm.org/D54911
llvm-svn: 350815
This modification is require for NetBSD with GCC, as there is a
custom unwind.h header implementation with different types.
No functional change intended for others.
Cherry-picked chunk from D33878.
llvm-svn: 347102
This modification is require for NetBSD with GCC, as there is a
custom unwind.h header implementation with different types.
No functional change intended for others.
Cherry-picked chunk from D33878.
llvm-svn: 347101
Summary:
Windows has always been LLP64, not LP64, so the macros were incorrect.
Check for _WIN64, since AArch64 sets that too.
The tests have to be fixed up in two main ways:
1. Use 'ULL' suffixes to avoid sign extension when passing hex literals
with the sign bit set to signed 128 arguments. Clang
-fms-compatibility makes such hex literals signed, not unsigned.
2. Disable various tests for 80-bit long double interoperation with i128
values.
Reviewers: joerg, mstorsjo
Subscribers: javed.absar, kristof.beyls, hiraditya, aheejin, llvm-commits
Differential Revision: https://reviews.llvm.org/D53918
llvm-svn: 345796
Summary:
In r81552, the HAS_80_BIT_LONG_DOUBLE macro was added to the unit test
only version of int_lib.h. One month later in r85260 the duplicate
int_lib.h was removed, but the tests still passed because we don't build
with -Werror.
This is the minimal change to bring it back, and I decided to put the
configuration macro next to our 128-bit integer support macro.
Reviewers: joerg, compnerd, mstorsjo
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D53838
llvm-svn: 345645
This breaks some configurations, so just forward declare everything that we need.
rdar://35943793
Differential revision: https://reviews.llvm.org/D50269
llvm-svn: 345551
Summary: This patch adds additional features and cpus from libgcc. Unfortunately we've overflowed the existing 32-bits of features so we had to add a new __cpu_features2 variable to hold the additional bits. This matches libgcc as far as I can tell.
Reviewers: echristo
Reviewed By: echristo
Subscribers: dberris, llvm-commits
Differential Revision: https://reviews.llvm.org/D53461
llvm-svn: 344830
This patch removes the instruction set check to make the msr APSR_nzcvq,
ip instruction only execute if Thumb2 is used.
The APSR is a subset of the bits of the CPSR
(B.1.3.3 of the Arm v7 A and R ARM [1]) and is only available for A and
R profiles.
However in section B.9.3.11 of the same document we see that:
"In the A and R profiles, APSR_nzcvq is the same as CPSR_f"
"ARM recommends the APSR forms when only the N, Z, C, V, Q, and GE[3:0]
bits are being written."
This patch also make those files assemble for Armv8-M Mainline
architecture profile.
The builtins were cross-compiled for Arm, Aarch64 and Armv6-M, Armv7-M
and Armv7E-M targets.
Cross-compiled tests were executed for Arm target.
[1]: https://developer.arm.com/docs/ddi0406/latest/arm-architecture-reference-manual-armv7-a-and-armv7-r-edition
Patch by hug-dev (Hugues de Valon).
Differential Revision: https://reviews.llvm.org/D51854
llvm-svn: 343601
Summary:
The complex division builtins (div?c3) use logb methods from libm to scale numbers during division and avoid rounding issues. However, these come from libm, meaning anyone that uses --rtlib=compiler-rt also has to include -lm. Implement logb* methods for standard ieee 754 floats so we can avoid -lm on those platforms, falling back to the old behavior (using either logb() or `__builtin_logb()`) when not supported.
These new methods are defined internally as `__compiler_rt_logb` so as not to conflict with the libm definitions in any way.
This fixes just the libm methods mentioned in PR32279 and PR28652. libc is still required, although that seems to not be an issue.
Note: this is proposed as an alternative to just adding -lm: D49330.
Reviewers: efriedma, compnerd, scanon, echristo
Reviewed By: echristo
Subscribers: jsji, echristo, nemanjai, dberris, mgorny, kbarton, delcypher, llvm-commits, #sanitizers
Differential Revision: https://reviews.llvm.org/D49514
llvm-svn: 342917
This is called by Bionic on dlclose to delete the emutls pthread key.
The return value of pthread_key_delete is unchecked and behaviour of
multiple calls to the method is dependent on the implementation of
pthread_key_delete.
Differential Revision: https://reviews.llvm.org/D52251
llvm-svn: 342608
On sparc64/Linux, sparc64 isn't defined; the canonical way of
checking for sparc64 is sparc && arch64, which also works on the
BSDs and Solaris. Since this problem does not occur on 32-bit
architectures, riscv32 can be ignored. This fixes and refines rL324593.
Patch by jrtc27 (James Clarke)
Differential Revision: https://reviews.llvm.org/D43146
llvm-svn: 342504
MachO symbols are prefixed with an extra '_' (that's 3 in total for this
function), so assembly calls have to go through a wrapper to insert any prefix
needed.
llvm-svn: 341540
Summary:
rL325492 disables FPU features when using soft floating point
(-mfloat-abi=soft), which is used internally when building for arm. This causes
errors with builtins that utililize VFP instructions.
With this change we check if VFP is enabled (by checking if the preprocessor
macro __VFP_FP__ is defined), and exclude such builtins if it is not enabled.
Reviewers: rengolin, samsonov, compnerd, smeenai, javed.absar, peter.smith
Reviewed By: peter.smith
Subscribers: delcypher, peter.smith, mgorny, kristof.beyls, chrib, llvm-commits
Differential Revision: https://reviews.llvm.org/D47217
llvm-svn: 338284
This function is available for linking in from kernel32.dll, but
it's not allowed to link that function from there in Windows Store
apps.
Differential Revision: https://reviews.llvm.org/D49055
llvm-svn: 337313
The initial implementaiton was using the C++ typeof keyword. This
causes the compiler to spew warnings unnecissarilly. This patch removes
the uses of typeof and replaces them with explicit typedefs of the
function types.
llvm-svn: 334981
The locks need to be implemented in a shared library to work correctly,
so they shouldn't be part of libclang_rt.builtins.a, except in
specialized scenarios where the user can prove it will only be linked
once.
Differential Revision: https://reviews.llvm.org/D47606
llvm-svn: 334779
Summary:
With Android/Bionic, delay deallocation to round 2 of 4. It must run after
C++ thread_local destructors have been called, but before the final 2
rounds, because emutls calls free, and jemalloc then needs another 2
rounds to free its thread-specific data.
Fixes https://github.com/android-ndk/ndk/issues/687
Reviewers: cmtice, srhines, jyknight, chh, echristo
Reviewed By: srhines, chh, echristo
Subscribers: echristo, delcypher, llvm-commits, #sanitizers
Differential Revision: https://reviews.llvm.org/D46978
llvm-svn: 334463
Summary:
rL325492 disables FPU features when using soft floating point
(-mfloat-abi=soft), which is used internally when building for armv7. This
causes errors with builtins that utililize VFP instructions. With this change
we first check if VFP is enabled (by checking if the preprocessor macro
__VFP_FP__ is defined) before including such builtins.
Reviewers: rengolin, samsonov, compnerd, smeenai, javed.absar, peter.smith
Reviewed By: peter.smith
Subscribers: peter.smith, mgorny, kristof.beyls, chrib, llvm-commits
Differential Revision: https://reviews.llvm.org/D47217
llvm-svn: 333216
Summary:
For RISCV32, we must force enable int128 for compiling long double routines using the flag -fforce-enable-int128.
Related clang patch: https://reviews.llvm.org/D43105
Reviewers: asb, kito-cheng, apazos, compnerd, howard.hinnant
Reviewed By: kito-cheng
Subscribers: shiva0217, efriedma, mgorny, hintonda, llvm-commits, #sanitizers
Differential Revision: https://reviews.llvm.org/D43106
llvm-svn: 326346
gcc resolves `__builtin_c?z` to `__c?zdi2` which leads to infinite
recursion. This problem has been observed for sparc64, mips64 and riscv.
Presumably this happens whenever an arch without dedicated bit counting
instructions is targeted. This patch provides a workaround.
Differential revision: https://reviews.llvm.org/D42902
llvm-svn: 324593
Generalize this handling to a separate toplevel ifdef (since any
windows case should use the same function), instead of indenting
the aarch64 case one step further.
Differential Revision: https://reviews.llvm.org/D42197
llvm-svn: 322928
As reported in PR35739, rL252927 added the Windows specific chkstk and
chkstk2 sources unconditionally, and since these are assembly without a
NO_EXEC_STACK_DIRECTIVE at the end, automated vulnerability scanners
warned about the objects having an executable stack.
Avoid the problem by only including these files when Windows is
targeted.
Reviewers: compnerd, rnk, martell
Reviewed By: martell
Subscribers: mstorsjo, mgorny, martell, javed.absar, #sanitizers, llvm-commits
Differential Revision: https://reviews.llvm.org/D41567
llvm-svn: 321431
Now that we have disabled the run-forever tests, and cleaned up the
intel 80-bit float based tests, we should be able to enable testing
compiler-rt for powerpc64.
llvm-svn: 319474
Summary:
Currently sys/cachectl.h is used unconditionally on MIPS although it is
only available on Linux and will fail the build when targeting baremetal
Reviewers: petarj
Reviewed By: petarj
Subscribers: sdardis, krytarowski
Differential Revision: https://reviews.llvm.org/D40659
llvm-svn: 319455
It is included in the built sources for all other arches supported
for MinGW currently, except for arm.
Differential Revision: https://reviews.llvm.org/D39938
llvm-svn: 318139
Summary:
The `b` instruction in Thumb1 has limited range, which may cause link-time errors if the jump target is far away.
This patch guards the tailcalls for non-Thumb1
Reviewers: peter.smith, compnerd, rengolin, eli.friedman
Reviewed By: rengolin
Subscribers: joerg, dalias, javed.absar, llvm-commits
Differential Revision: https://reviews.llvm.org/D39700
llvm-svn: 317814
r303188 removed all the uses of aliases for EABI functions from
compiler-rt, because some of them had mismatched calling conventions.
Obviously, we can't use aliases for functions which don't have the same
calling convention, but that's only an issue for floating-point
functions with the hardfloat ABI. In other cases, the stubs increase
size and reduce performance for no benefit.
This patch adds back the aliases, with appropriate checks to make sure
they're only used in cases where the calling convention matches.
llvm-svn: 314851
Summary:
clang does not assemble files in thumb mode unless .thumb declaration
is present. Add .thumb/.arm decl to _FUNCTION macros to ensure that
files are assembled correctly.
Also add a fix to ensure that armv7k-watchos can assemble the
aeabi_c{f|d}cmp.S files.
Fixes PR 34715.
Reviewers: compnerd, peter.smith, srhines, weimingz, rengolin, efriedma, t.p.northover, fjricci
Reviewed By: compnerd
Subscribers: aemerson, javed.absar, llvm-commits, kristof.beyls
Differential Revision: https://reviews.llvm.org/D38390
llvm-svn: 314718
Lei Huang