__start_/__stop_ references retain C identifier name sections such as
__llvm_prf_*. Putting these into a section group disables this logic.
The ELF section group semantics ensures that group members are retained
or discarded as a unit. When a function symbol is discarded, this allows
allows linker to discard counters, data and values associated with that
function symbol as well.
Note that `noduplicates` COMDAT is lowered to zero-flag section group in
ELF. We only set this for functions that aren't already in a COMDAT and
for those that don't have available_externally linkage since we already
use regular COMDAT groups for those.
Differential Revision: https://reviews.llvm.org/D96757
C identifier name input sections such as __llvm_prf_* are GC roots so
they cannot be discarded. In LLD, the SHF_LINK_ORDER flag overrides the
C identifier name semantics.
The !associated metadata may be attached to a global object declaration
with a single argument that references another global object, and it
gets lowered to SHF_LINK_ORDER flag. When a function symbol is discarded
by the linker, setting up !associated metadata allows linker to discard
counters, data and values associated with that function symbol.
Note that !associated metadata is only supported by ELF, it does not have
any effect on non-ELF targets.
Differential Revision: https://reviews.llvm.org/D76802
C identifier name input sections such as __llvm_prf_* are GC roots so
they cannot be discarded. In LLD, the SHF_LINK_ORDER flag overrides the
C identifier name semantics.
The !associated metadata may be attached to a global object declaration
with a single argument that references another global object, and it
gets lowered to SHF_LINK_ORDER flag. When a function symbol is discarded
by the linker, setting up !associated metadata allows linker to discard
counters, data and values associated with that function symbol.
Note that !associated metadata is only supported by ELF, it does not have
any effect on non-ELF targets.
Differential Revision: https://reviews.llvm.org/D76802
llvm-cov -path-equivalence=/tmp,... is used by some checked-in coverage mapping
files where the original filename is under /tmp. If the test itself produces the
coverage mapping file, there is no need for /tmp.
For coverage_emptylines.cpp: the source filename is under the build directory.
If the build directory is under /tmp, the path mapping will make
llvm-cov fail to find the file.
CallInst::updateProfWeight() creates branch_weights with i64 instead of i32.
To be more consistent everywhere and remove lots of casts from uint64_t
to uint32_t, use i64 for branch_weights.
Reviewed By: davidxl
Differential Revision: https://reviews.llvm.org/D88609
CallInst::updateProfWeight() creates branch_weights with i64 instead of i32.
To be more consistent everywhere and remove lots of casts from uint64_t
to uint32_t, use i64 for branch_weights.
Reviewed By: davidxl
Differential Revision: https://reviews.llvm.org/D88609
Add support for expanding the %t filename specifier in LLVM_PROFILE_FILE
to the TMPDIR environment variable. This is supported on all platforms.
On Darwin, TMPDIR is used to specify a temporary application-specific
scratch directory. When testing apps on remote devices, it can be
challenging for the host device to determine the correct TMPDIR, so it's
helpful to have the runtime do this work.
rdar://68524185
Differential Revision: https://reviews.llvm.org/D87332
gcov is an "Edge Profiling with Edge Counters" application according to
Optimally Profiling and Tracing Programs (1994).
The minimum number of counters necessary is |E|-(|V|-1). The unmeasured edges
form a spanning tree. Both GCC --coverage and clang -fprofile-generate leverage
this optimization. This patch implements the optimization for clang --coverage.
The produced .gcda files are much smaller now.
As reported in Bug 42535, `clang` doesn't inline atomic ops on 32-bit
Sparc, unlike `gcc` on Solaris. In a 1-stage build with `gcc`, only two
testcases are affected (currently `XFAIL`ed), while in a 2-stage build more
than 100 tests `FAIL` due to this issue.
The reason for this `gcc`/`clang` difference is that `gcc` on 32-bit
Solaris/SPARC defaults to `-mpcu=v9` where atomic ops are supported, unlike
with `clang`'s default of `-mcpu=v8`. This patch changes `clang` to use
`-mcpu=v9` on 32-bit Solaris/SPARC, too.
Doing so uncovered two bugs:
`clang -m32 -mcpu=v9` chokes with any Solaris system headers included:
/usr/include/sys/isa_defs.h:461:2: error: "Both _ILP32 and _LP64 are defined"
#error "Both _ILP32 and _LP64 are defined"
While `clang` currently defines `__sparcv9` in a 32-bit `-mcpu=v9`
compilation, neither `gcc` nor Studio `cc` do. In fact, the Studio 12.6
`cc(1)` man page clearly states:
These predefinitions are valid in all modes:
[...]
__sparcv8 (SPARC)
__sparcv9 (SPARC -m64)
At the same time, the patch defines `__GCC_HAVE_SYNC_COMPARE_AND_SWAP_[1248]`
for a 32-bit Sparc compilation with any V9 cpu. I've also changed
`MaxAtomicInlineWidth` for V9, matching what `gcc` does and the Oracle
Developer Studio 12.6: C User's Guide documents (Ch. 3, Support for Atomic
Types, 3.1 Size and Alignment of Atomic C Types).
The two testcases that had been `XFAIL`ed for Bug 42535 are un-`XFAIL`ed
again.
Tested on `sparcv9-sun-solaris2.11` and `amd64-pc-solaris2.11`.
Differential Revision: https://reviews.llvm.org/D86621
For a CFG G=(V,E), Knuth describes that by Kirchoff's circuit law, the minimum
number of counters necessary is |E|-(|V|-1). The emitted edges form a spanning
tree. libgcov emitted .gcda files leverages this optimization while clang
--coverage's doesn't.
Propagate counts by Kirchhoff's circuit law so that llvm-cov gcov can
correctly print line counts of gcc --coverage emitted files and enable
the future improvement of clang --coverage.
Two tests `FAIL` on 32-bit sparc:
Profile-sparc :: Posix/instrprof-gcov-parallel.test
UBSan-Standalone-sparc :: TestCases/Float/cast-overflow.cpp
The failure mode is similar:
Undefined first referenced
symbol in file
__atomic_store_4 /var/tmp/instrprof-gcov-parallel-6afe8d.o
__atomic_load_4 /var/tmp/instrprof-gcov-parallel-6afe8d.o
Undefined first referenced
symbol in file
__atomic_load_1 /var/tmp/cast-overflow-72a808.o
This is a known bug: `clang` doesn't inline atomics on 32-bit sparc, unlike
`gcc`.
The patch therefore `XFAIL`s the tests.
Tested on `sparcv9-sun-solaris2.11` and `amd64-pc-solaris2.11`.
Differential Revision: https://reviews.llvm.org/D85346
Similarly as for pointers, even for integers a == b is usually false.
GCC also uses this heuristic.
Reviewed By: ebrevnov
Differential Revision: https://reviews.llvm.org/D85781
InstrProfilingBuffer.c.o is generic code that must support compilation
into freestanding projects. This gets rid of its dependence on the
_getpagesize symbol from libc, shifting it to InstrProfilingFile.c.o.
This fixes a build failure seen in a firmware project.
rdar://66249701
Otherwise if 'ld' is an older system LLD (FreeBSD; or if someone adds 'ld' to
point to an LLD from a different installation) which does not support the
current ModuleSummaryIndex::BitCodeSummaryVersion, the test will fail.
Add lit feature 'binutils_lto'. GNU ld is more common than GNU gold, so
we can just require 'is_binutils_lto_supported' to additionally support GNU ld.
Reviewed By: myhsu
Differential Revision: https://reviews.llvm.org/D84133
... on systems where wait() isn't one of the declarations transitively included
via unistd.h (i.e. Darwin).
Differential Revision: https://reviews.llvm.org/D84207
GCC r187297 (2012-05) introduced `__gcov_dump` and `__gcov_reset`.
`__gcov_flush = __gcov_dump + __gcov_reset`
The resolution to https://gcc.gnu.org/PR93623 ("No need to dump gcdas when forking" target GCC 11.0) removed the unuseful and undocumented __gcov_flush.
Close PR38064.
Reviewed By: calixte, serge-sans-paille
Differential Revision: https://reviews.llvm.org/D83149
Similar to the reason behind moving __llvm_profile_filename into a
separate file[1]. When users try to use Full LTO with BFD linker to
generate IR level PGO profile, the __llvm_profile_raw_version variable,
which is used for marking instrumentation level, generated by frontend
would somehow conflict with the weak symbol provided by profiling
runtime.
In most of the cases, BFD linkers will pick profiling runtime's weak symbol
as the real definition and thus generate the incorrect instrumentation
level metadata in the final executables.
Moving __llvm_profile_raw_version into a separate file would make
linkers not seeing the weak symbol in the archive unless the frontend
doesn't generate one.
[1] https://reviews.llvm.org/D34797
Differential Revision: https://reviews.llvm.org/D83967
This test spawns 32 child processes which race to update counters on
shared memory pages. On some Apple-internal machines, two processes race
to perform an update in approximately 0.5% of the test runs, leading to
dropped counter updates. Deflake the test by using atomic increments.
Tested with:
```
$ for I in $(seq 1 1000); do echo ":: Test run $I..."; ./bin/llvm-lit projects/compiler-rt/test/profile/Profile-x86_64h/ContinuousSyncMode/online-merging.c -av || break; done
```
rdar://64956774
atexit registered functions run earlier so `__attribute__((destructor))`
annotated functions cannot be tracked.
Set a priority of 100 (compatible with GCC 7 onwards) to track
destructors and destructors whose priorities are greater than 100.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=7970
Reviewed By: calixte, marco-c
Differential Revision: https://reviews.llvm.org/D82253
The !associated metadata may be attached to a global object declaration
with a single argument that references another global object. This
metadata prevents discarding of the global object in linker GC unless
the referenced object is also discarded.
Furthermore, when a function symbol is discarded by the linker, setting
up !associated metadata allows linker to discard counters, data and
values associated with that function symbol. This is not possible today
because there's metadata to guide the linker. This approach is also used
by other instrumentations like sanitizers.
Note that !associated metadata is only supported by ELF, it does not have
any effect on non-ELF targets.
Differential Revision: https://reviews.llvm.org/D76802
The !associated metadata may be attached to a global object declaration
with a single argument that references another global object. This
metadata prevents discarding of the global object in linker GC unless
the referenced object is also discarded.
Furthermore, when a function symbol is discarded by the linker, setting
up !associated metadata allows linker to discard counters, data and
values associated with that function symbol. This is not possible today
because there's metadata to guide the linker. This approach is also used
by other instrumentations like sanitizers.
Note that !associated metadata is only supported by ELF, it does not have
any effect on non-ELF targets.
Differential Revision: https://reviews.llvm.org/D76802
It seems that after dc52ce424b, all big-endian problems have been fixed.
01899bb4e4 seems to have fixed XFAIL: * of
profile/instrprof-gcov-__gcov_flush-terminate.test
This essentially reverts commit 5a9b792d72 and
93d5ae3af1.
global-ctor.ll no longer checks what it intended to check
(@_GLOBAL__sub_I_global-ctor.ll needs a !dbg to work).
Rewrite it.
gcov 3.4 and gcov 4.2 use the same format, thus we can lower the version
requirement to 3.4
Fixes PR45673
The commit 9180c14fe4 (D76206) resolved only a part of the problem
of concurrent .gcda file creation. It ensured that only one process
creates the file but did not ensure that the process locks the
file first. If not, the process which created the file may clobber
the contents written by a process which locked the file first.
This is the cause of PR45673.
This commit prevents the clobbering by revising the assumption
that a process which creates the file locks the file first.
Regardless of file creation, a process which locked the file first
uses fwrite (new_file==1) and other processes use mmap (new_file==0).
I also tried to keep the creation/first-lock process same by using
mkstemp/link/unlink but the code gets long. This commit is more
simple.
Note: You may be confused with other changes which try to resolve
concurrent file access. My understanding is (may not be correct):
D76206: Resolve race of .gcda file creation (but not lock)
This one: Resolve race of .gcda file creation and lock
D54599: Same as D76206 but abandoned?
D70910: Resolve race of multi-threaded counter flushing
D74953: Resolve counter sharing between parent/children processes
D78477: Revision of D74953
Differential Revision: https://reviews.llvm.org/D79556
Summary:
When forking in several threads, the counters were written out in using the same global static variables (see GCDAProfiling.c): that leads to crashes.
So when there is a fork, the counters are resetted in the child process and they will be dumped at exit using the interprocess file locking.
When there is an exec, the counters are written out and in case of failures they're resetted.
Reviewers: jfb, vsk, marco-c, serge-sans-paille
Reviewed By: marco-c, serge-sans-paille
Subscribers: llvm-commits, serge-sans-paille, dmajor, cfe-commits, hiraditya, dexonsmith, #sanitizers, marco-c, sylvestre.ledru
Tags: #sanitizers, #clang, #llvm
Differential Revision: https://reviews.llvm.org/D78477
The intent of the `llvm_gcda_start_file` function is that only
one process create the .gcda file and initialize it to be updated
by other processes later.
Before this change, if multiple processes are started simultaneously,
some of them may initialize the file because both the first and
second `open` calls may succeed in a race condition and `new_file`
becomes 1 in those processes. This leads incorrect coverage counter
values. This often happens in MPI (Message Passing Interface) programs.
The test program added in this change is a simple reproducer.
This change ensures only one process creates/initializes the file by
using the `O_EXCL` flag.
Differential Revision: https://reviews.llvm.org/D76206
After the format change from D69471, there can be more than one section
in an object that contains coverage function records. Look up each of
these sections and concatenate all the records together.
This re-enables the instrprof-merging.cpp test, which previously was
failing on OSes which use comdats.
Thanks to Jeremy Morse, who very kindly provided object files from the
bot I broke to help me debug.
An execution count goes missing for a constructor, this needs
investigation:
http://lab.llvm.org:8011/builders/clang-ppc64be-linux/builds/45132/
```
/home/buildbots/ppc64be-clang-test/clang-ppc64be/llvm/compiler-rt/test/profile/instrprof-merging.cpp:28:16:
error: V1: expected string not found in input
A() {} // V1: [[@LINE]]{{ *}}|{{ *}}1
<stdin>:28:32: note: possible intended match here
28| | A() {} // V1: [[@LINE]]{{ *}}|{{ *}}1
```
Summary: `.cmd` is interpreted as script in windows console.
Reviewers: davidxl, rnk
Reviewed By: davidxl
Differential Revision: https://reviews.llvm.org/D73327
This is an alternative to the continous mode that was implemented in
D68351. This mode relies on padding and the ability to mmap a file over
the existing mapping which is generally only available on POSIX systems
and isn't suitable for other platforms.
This change instead introduces the ability to relocate counters at
runtime using a level of indirection. On every counter access, we add a
bias to the counter address. This bias is stored in a symbol that's
provided by the profile runtime and is initially set to zero, meaning no
relocation. The runtime can mmap the profile into memory at abitrary
location, and set bias to the offset between the original and the new
counter location, at which point every subsequent counter access will be
to the new location, which allows updating profile directly akin to the
continous mode.
The advantage of this implementation is that doesn't require any special
OS support. The disadvantage is the extra overhead due to additional
instructions required for each counter access (overhead both in terms of
binary size and performance) plus duplication of counters (i.e. one copy
in the binary itself and another copy that's mmapped).
Differential Revision: https://reviews.llvm.org/D69740
The executable acquires an advisory record lock (`fcntl(fd, F_SETLKW, *)`) on a profile file.
Merge pool size >= 10 may be beneficial when the concurrency is large.
Also fix a small problem about snprintf. It can cause the filename to be truncated after %m.
Reviewed By: davidxl
Differential Revision: https://reviews.llvm.org/D71970
See PR43425:
https://bugs.llvm.org/show_bug.cgi?id=43425
When writing profile data on Windows we were opening profile file with
exclusive read/write access.
In case we are trying to write to the file from multiple processes
simultaneously, subsequent calls to CreateFileA would return
INVALID_HANDLE_VALUE.
To fix this, I changed to open without exclusive access and then take a
lock.
Patch by Michael Holman!
Differential revision: https://reviews.llvm.org/D70330
Make it possible to use online profile merging ("%m" mode) with
continuous sync ("%c" mode).
To implement this, the merged profile is locked in the runtime
initialization step and either a) filled out for the first time or b)
checked for compatibility. Then, the profile can simply be mmap()'d with
MAP_SHARED set. With the mmap() in place, counter updates from every
process which uses an image are mapped onto the same set of physical
pages assigned by the filesystem cache. After the mmap() is set up, the
profile is unlocked.
Differential Revision: https://reviews.llvm.org/D69586
Add support for continuously syncing profile counter updates to a file.
The motivation for this is that programs do not always exit cleanly. On
iOS, for example, programs are usually killed via a signal from the OS.
Running atexit() handlers after catching a signal is unreliable, so some
method for progressively writing out profile data is necessary.
The approach taken here is to mmap() the `__llvm_prf_cnts` section onto
a raw profile. To do this, the linker must page-align the counter and
data sections, and the runtime must ensure that counters are mapped to a
page-aligned offset within a raw profile.
Continuous mode is (for the moment) incompatible with the online merging
mode. This limitation is lifted in https://reviews.llvm.org/D69586.
Continuous mode is also (for the moment) incompatible with value
profiling, as I'm not sure whether there is interest in this and the
implementation may be tricky.
As I have not been able to test extensively on non-Darwin platforms,
only Darwin support is included for the moment. However, continuous mode
may "just work" without modification on Linux and some UNIX-likes. AIUI
the default value for the GNU linker's `--section-alignment` flag is set
to the page size on many systems. This appears to be true for LLD as
well, as its `no_nmagic` option is on by default. Continuous mode will
not "just work" on Fuchsia or Windows, as it's not possible to mmap() a
section on these platforms. There is a proposal to add a layer of
indirection to the profile instrumentation to support these platforms.
rdar://54210980
Differential Revision: https://reviews.llvm.org/D68351
When the %m filename pattern is used, the filename is unique to each
image, so the cached value is wrong.
It struck me that the full filename isn't something that's recomputed
often, so perhaps it doesn't need to be cached at all. David Li pointed
out we can go further and just hide lprofCurFilename. This may regress
workflows that depend on using the set-filename API to change filenames
across all loaded DSOs, but this is expected to be very rare.
rdar://55137071
Differential Revision: https://reviews.llvm.org/D69137
llvm-svn: 375301
This test remains flaky everywhere, I think. We should consider deleting
it and accompanying support code in GCOVProfiling: I've stopped short of
doing that now as the gcov exec* tests appear to be stable.
See the thread re: r347779.
llvm-svn: 373121
This fixes a test failure in instrprof-set-file-object-merging.c which
seems to have been caused by reuse of stale data in old raw profiles.
llvm-svn: 372041
While working on https://reviews.llvm.org/D40900 (which effectively is about enabling compiler-rt on sparc these days), I came across two failing profile testcases:
Profile-sparc :: instrprof-merge-match.test
Profile-sparc :: instrprof-merge.c
Profile-sparcv9 :: instrprof-merge-match.test
Profile-sparcv9 :: instrprof-merge.c
All of them crashed with a SIGBUS in __llvm_profile_merge_from_buffer:
Thread 2 received signal SIGSEGV, Segmentation fault.
[Switching to Thread 1 (LWP 1)]
0x00012368 in __llvm_profile_merge_from_buffer (
ProfileData=0x2384c <main.Buffer> "\377lprofR\201", ProfileSize=360)
at /vol/llvm/src/llvm/local/projects/compiler-rt/lib/profile/InstrProfilingMerge.c:95
95 SrcDataEnd = SrcDataStart + Header->DataSize;
where Header is insufficiently aligned for a strict-alignment target like SPARC.
Fixed by forcing the alignment to uint64_t, the members of struct __llvm_profile_header,
in the callers.
Tested on sparcv9-sun-solaris2.11.
https://reviews.llvm.org/D64498
llvm-svn: 365805
These lit configuration files are really Python source code. Using the
.py file extension helps editors and tools use the correct language
mode. LLVM and Clang already use this convention for lit configuration,
this change simply applies it to all of compiler-rt.
Reviewers: vitalybuka, dberris
Differential Revision: https://reviews.llvm.org/D63658
llvm-svn: 364591
Summary:
User code can open a file on its own and pass it to the runtime, rather than
specifying a name and having the runtime open the file. This supports the use
case where a process cannot open a file on its own but can receive a file
descriptor from another process.
Relanding https://reviews.llvm.org/D62541. The original revision unlocked
the file before calling flush, this revision fixes that.
Reviewers: Dor1s, davidxl
Reviewed By: Dor1s
Subscribers: #sanitizers, llvm-commits
Tags: #sanitizers, #llvm
Differential Revision: https://reviews.llvm.org/D63581
llvm-svn: 364231
This caused instrumented Clang to become crashy. See llvm-commits thread
for repro steps.
This also reverts follow-up r362716 which added test cases.
> Author: Sajjad Mirza
>
> Differential Revision: http://reviews.llvm.org/D62541
llvm-svn: 363134
Patch from 'troyj':
Hi, I ran into a problem with this test when the source was located in
certain directories. The mkdir(2) man page states that the set-group-ID
bit is inherited from the parent directory, but this test was written in
such a way that it assumes the bit is unset. Whether that assumption is
true depends on where the checkout lives, which leads to some people
being able to reproduce the problem whereas others cannot. I think the
correct fix is to exclude the bit from the check.
Making probinson a reviewer since they reviewed the original test.
Patch landed for troyj, thanks!
Differential Revision: D53832
llvm-svn: 357449
r355343 was landed and was reverted in r355363 due to build breakage.
This patch adds Linux/Windows support on top of r355343.
In this patch, Darwin should be working with testing case. Linux should be working,
I will enable the testing case in a follwup diff. Windows/Other should be building.
Correct implementation for Other platforms will be added.
Thanks David for reviewing the original diff, helping me with issues on Linux, and
giving suggestions for adding support for Other platforms.
llvm-svn: 355701
Summary:
I hadn't realized that instrumentation runs before inlining, so we can't
use the function as the comdat group. Doing so can create relocations
against discarded sections when references to discarded __profc_
variables are inlined into functions outside the function's comdat
group.
In the future, perhaps we should consider standardizing the comdat group
names that ELF and COFF use. It will save object file size, since
__profv_$sym won't appear in the symbol table again.
Reviewers: xur, vsk
Subscribers: eraman, hiraditya, cfe-commits, #sanitizers, llvm-commits
Tags: #clang, #sanitizers, #llvm
Differential Revision: https://reviews.llvm.org/D58737
llvm-svn: 355044
Summary:
The llvm-cov tool needs to be able to find coverage names in the
executable, so the .lprfn and .lcovmap sections cannot be merged into
.rdata.
Also, the linker merges .lprfn$M into .lprfn, so llvm-cov needs to
handle that when looking up sections. It has to support running on both
relocatable object files and linked PE files.
Lastly, when loading .lprfn from a PE file, llvm-cov needs to skip the
leading zero byte added by the profile runtime.
Reviewers: vsk
Subscribers: hiraditya, #sanitizers, llvm-commits
Tags: #sanitizers, #llvm
Differential Revision: https://reviews.llvm.org/D58661
llvm-svn: 354840
Summary:
The changes in https://reviews.llvm.org/D44847 cause load time failure
due to lprofMergeValueProfData in Android libs enabled with profile generation:
"dlopen failed: cannot locate symbol "lprofMergeValueProfData" referenced by..."
Marking lprofMergeValueProfData as hidden so the correct in-module definition
is picked by the linker.
Reviewers: davidxl
Reviewed By: davidxl
Subscribers: efriedma, xur, davidxl, llvm-commits
Differential Revision: https://reviews.llvm.org/D55893
llvm-svn: 354064
Petr Hosek