<html>
<head>
<base href="http://llvm.org/bugs/" />
</head>
<body><table border="1" cellspacing="0" cellpadding="8">
<tr>
<th>Bug ID</th>
<td><a class="bz_bug_link
bz_status_NEW "
title="NEW --- - [powerpc-darwin8] missing 64b atomic operations on PPC32"
href="http://llvm.org/bugs/show_bug.cgi?id=18470">18470</a>
</td>
</tr>
<tr>
<th>Summary</th>
<td>[powerpc-darwin8] missing 64b atomic operations on PPC32
</td>
</tr>
<tr>
<th>Product</th>
<td>compiler-rt
</td>
</tr>
<tr>
<th>Version</th>
<td>unspecified
</td>
</tr>
<tr>
<th>Hardware</th>
<td>Macintosh
</td>
</tr>
<tr>
<th>OS</th>
<td>MacOS X
</td>
</tr>
<tr>
<th>Status</th>
<td>NEW
</td>
</tr>
<tr>
<th>Severity</th>
<td>normal
</td>
</tr>
<tr>
<th>Priority</th>
<td>P
</td>
</tr>
<tr>
<th>Component</th>
<td>compiler-rt
</td>
</tr>
<tr>
<th>Assignee</th>
<td>unassignedbugs@nondot.org
</td>
</tr>
<tr>
<th>Reporter</th>
<td>fang@csl.cornell.edu
</td>
</tr>
<tr>
<th>CC</th>
<td>llvmbugs@cs.uiuc.edu
</td>
</tr>
<tr>
<th>Classification</th>
<td>Unclassified
</td>
</tr></table>
<p>
<div>
<pre>I noticed during libc++ tests on powerpc-darwin8 that some tests fail due to
lack of 64b atomics (on PPC32).
In file included from integral.pass.cpp:87:
/Users/fang/local/src/LLVM-svn/libcxx.git/include/atomic:574:10: error: atomic
store requires runtime suppor
t that is not available for this target
{__c11_atomic_store(&__a_, __d, __m);}
^
integral.pass.cpp:102:9: note: in instantiation of member function
'std::__1::__atomic_base<long long, false
<span class="quote">>::store' requested here</span >
obj.store(T(0));
^
integral.pass.cpp:159:5: note: in instantiation of function template
specialization 'do_test<std::__1::atomi
c<long long>, long long>' requested here
do_test<A, T>();
^
integral.pass.cpp:175:5: note: in instantiation of function template
specialization 'test<std::__1::atomic<l
ong long>, long long>' requested here
test<std::atomic_llong, long long>();
^
Can a 32b architecture emulate 64b atomic ops?
On OS X, darwin8, <libkern/OSAtomic.h> provides prototypse for 64b atomics for
!PPC32:
#if defined(__ppc64__) || defined(__i386__) || defined(__x86_64__)
int64_t OSAtomicAdd64( int64_t theAmount, int64_t *theValue );
int64_t OSAtomicAdd64Barrier( int64_t theAmount, int64_t *theValue );
inline static
int64_t OSAtomicIncrement64( int64_t *theValue )
{ return OSAtomicAdd64( 1, theValue); }
inline static
int64_t OSAtomicIncrement64Barrier( int64_t *theValue )
{ return OSAtomicAdd64Barrier( 1, theValue); }
inline static
int64_t OSAtomicDecrement64( int64_t *theValue )
{ return OSAtomicAdd64( -1, theValue); }
inline static
int64_t OSAtomicDecrement64Barrier( int64_t *theValue )
{ return OSAtomicAdd64Barrier( -1, theValue); }
#endif /* defined(__ppc64__) || defined(__i386__) || defined(__x86_64__) */
so if it's possible, I might end up having to provide my own definitions.</pre>
</div>
</p>
<hr>
<span>You are receiving this mail because:</span>
<ul>
<li>You are on the CC list for the bug.</li>
</ul>
</body>
</html>