make check stops after one case in asan.exp
Hi,
I ran into a problem that gcc make check stops after one case in
asan.exp, here is the dump information:
Executed ./sleep-before-dying-1.exe, status 1
=
==30519== ERROR: AddressSanitizer: heap-use-after-free on address
0x41869fc5 at pc 0x80486dd bp 0xbf8f3728 sp 0xbf8f371c
READ of size 1 at 0x41869fc5 thread T0
#0 0x80486dc
(/home/.../build/gcc/testsuite/gcc/sleep-before-dying-1.exe+0x80486dc)
#1 0x41598b55 (/lib/tls/i686/cmov/libc-2.10.1.so+0x16b55)
#2 0x8048580
(/home/.../build/gcc/testsuite/gcc/sleep-before-dying-1.exe+0x8048580)
0x41869fc5 is located 5 bytes inside of 10-byte region [0x41869fc0,0x41869fca)
freed by thread T0 here:
#0 0x40069004
(/home/.../build/i686-pc-linux-gnu/libsanitizer/asan/.libs/libasan.so.0.0.0+0x10004)
#1 0x80486af
(/home/.../build/gcc/testsuite/gcc/sleep-before-dying-1.exe+0x80486af)
#2 0x41598b55 (/lib/tls/i686/cmov/libc-2.10.1.so+0x16b55)
#3 0x8048580
(/home/.../build/gcc/testsuite/gcc/sleep-before-dying-1.exe+0x8048580)
previously allocated by thread T0 here:
#0 0x400690e4
(/home/.../build/i686-pc-linux-gnu/libsanitizer/asan/.libs/libasan.so.0.0.0+0x100e4)
#1 0x80486a5
(/home/.../build/gcc/testsuite/gcc/sleep-before-dying-1.exe+0x80486a5)
#2 0x41598b55 (/lib/tls/i686/cmov/libc-2.10.1.so+0x16b55)
#3 0x8048580
(/home/.../build/gcc/testsuite/gcc/sleep-before-dying-1.exe+0x8048580)
Shadow bytes around the buggy address:
0x2830d3a0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x2830d3b0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x2830d3c0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x2830d3d0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x2830d3e0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
=>0x2830d3f0: fa fa fa fa fa fa fa fa[fd]fd fd fd fd fd fd fd
0x2830d400: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x2830d410: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x2830d420: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x2830d430: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x2830d440: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
Shadow byte legend (one shadow byte represents 8 application bytes):
Addressable: 00
Partially addressable: 01 02 03 04 05 06 07
Heap left redzone: fa
Heap righ redzone: fb
Freed Heap region: fd
Stack left redzone:f1
Stack mid redzone: f2
Stack right redzone: f3
Stack partial redzone: f4
Stack after return:f5
Stack use after scope: f8
Global redzone:f9
Global init order: f6
Poisoned by user: f7
ASan internal: fe
Stats: 0M malloced (0M for red zones) by 1 calls
Stats: 0M realloced by 0 calls
Stats: 0M freed by 1 calls
Stats: 0M really freed by 0 calls
Stats: 0M (0M-0M) mmaped; 1 maps, 0 unmaps
mmaps by size class: 7:4095;
mallocs by size class: 7:1;
frees by size class: 7:1;
rfrees by size class:
Stats: malloc large: 0 small slow: 1
Stats: StackDepot: 0 ids; 0M mapped
==30519== ABORTING
==30519== Sleeping for 1 second(s)
ERROR: (DejaGnu) proc "lreverse {{ASAN_OPTIONS 0}}" does not exist.
The error code is CHILDSTATUS 30519 1
The info on the error is:
child process exited abnormally
while executing
"close file8"
make check just exists after this error and I can only get partial
test results like:
=== g++ Summary ===
# of expected passes192
# of unsupported tests 9
So anything I missed? Thanks very much.
The gcc is configured with :
../gcc/configure
--prefix=/home/.../target/
--with-gmp=/home/.../usr/
--with-mpfr=/home/.../usr/
--with-mpc=/home/.../usr/
--with-ppl=/home/.../usr/
--with-cloog=/home/.../usr/
--with-host-libstdcxx='-static-libgcc
-Wl,-L/usr/lib/gcc/i486-linux-gnu/4.4.1/,-Bstatic,-lstdc++,-Bdynamic
-lm'
--enable-languages=c,c++,lto
The triplet should be i686-pc-linux-gnu by default.
--
Best Regards.
Re: make check stops after one case in asan.exp
"Bin.Cheng" writes:
> ERROR: (DejaGnu) proc "lreverse {{ASAN_OPTIONS 0}}" does not exist.
What is the version of tcl you are using? Perhaps it doesn't know about
lreverse yet (which was added in tcl-8.5).
Andreas.
--
Andreas Schwab, sch...@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756 01D3 44D5 214B 8276 4ED5
"And now for something completely different."
Re: make check stops after one case in asan.exp
On Sat, Jan 19, 2013 at 5:52 PM, Andreas Schwab wrote:
> "Bin.Cheng" writes:
>
>> ERROR: (DejaGnu) proc "lreverse {{ASAN_OPTIONS 0}}" does not exist.
>
> What is the version of tcl you are using? Perhaps it doesn't know about
> lreverse yet (which was added in tcl-8.5).
My version is tcl-8.4, I will update it and try again.
Thanks for helping.
--
Best Regards.
gcc 4.7 regression on bool function specialization
/*
*
* The output of this with gcc 4.7.2 is:
*
* 1
* 2
* this generic-int function should not be called
* 0
*
*
*
* The output of this with gcc 4.6.3 is:
*
* 1
* 2
* this type-bool function SHOULD BE called
* 1
*
*/
#include
#include
#include
#include
struct Type { int dummy; };
template
inline T extractTag(const M &map, const char *key)
{
if (!strcmp(key, "blah-bool"))
std::cout << "this generic-int function should not be called"
<< std::endl;
T ret = 0;
std::istringstream stream(extractTag(map, key));
stream >> ret;
return ret;
}
template
inline bool extractTag(const M &map, const char *key)
{
if (!strcmp(key, "blah-bool"))
std::cout << "this generic-bool function should not be called"
<< std::endl;
std::string str(extractTag(map, key));
std::transform(str.begin(), str.end(), str.begin(), ::tolower);
return (str == "1" || str == "true");
}
template <>
inline unsigned int extractTag(const Type&, const char *key)
{
return 2;
}
template <>
inline std::string extractTag(const Type&, const char *key)
{
if (!strcmp(key, "blah-int"))
return "1";
if (!strcmp(key, "blah-unsigned-int"))
return "-should-not-be-reached-";
if (!strcmp(key, "blah-bool"))
return "true";
return "invalid";
}
template <>
inline bool extractTag(const Type&, const char *key)
{
if (!strcmp(key, "blah-bool"))
std::cout << "this type-bool function SHOULD BE called" <<
std::endl;
return true;
}
int main(int argc, char *argv[])
{
Type t;
std::cout << extractTag(t, "blah-int") << std::endl;
std::cout << extractTag(t, "blah-unsigned-int") <<
std::endl;
std::cout << extractTag(t, "blah-bool") << std::endl;
return 0;
}
Re: gcc 4.7 regression on bool function specialization
N.B. This mail is not appropriate on this mailing list, which is for
discussion of development of GCC. For help with GCC use the gcc-help
list or to report bugs use bugzilla, thanks.
On 19 January 2013 18:58, Jason A. Donenfeld wrote:
> /*
> *
> * The output of this with gcc 4.7.2 is:
> *
> * 1
> * 2
> * this generic-int function should not be called
> * 0
This is the correct behaviour, the old behaviour was a bug.
> struct Type { int dummy; };
>
> template
> inline T extractTag(const M &map, const char *key)
> {
> if (!strcmp(key, "blah-bool"))
> std::cout << "this generic-int function should not be called"
> << std::endl;
> T ret = 0;
> std::istringstream stream(extractTag(map, key));
> stream >> ret;
> return ret;
> }
> template
> inline bool extractTag(const M &map, const char *key)
> {
> if (!strcmp(key, "blah-bool"))
> std::cout << "this generic-bool function should not be
> called" << std::endl;
> std::string str(extractTag(map, key));
> std::transform(str.begin(), str.end(), str.begin(), ::tolower);
> return (str == "1" || str == "true");
> }
This is not a partial specialization (you cannot have partial
specializations of function templates only class templates)so this is
a new overload of extractTag with a single template parameter.
> template <>
> inline bool extractTag(const Type&, const char *key)
> {
> if (!strcmp(key, "blah-bool"))
> std::cout << "this type-bool function SHOULD BE called" <<
> std::endl;
> return true;
> }
This is a specialization of extractTag not extractTag, so
cannot be called by extractTag(t, "blah-bool") because argument
deduction fails.
gcc-4.7-20130119 is now available
Snapshot gcc-4.7-20130119 is now available on ftp://gcc.gnu.org/pub/gcc/snapshots/4.7-20130119/ and on various mirrors, see http://gcc.gnu.org/mirrors.html for details. This snapshot has been generated from the GCC 4.7 SVN branch with the following options: svn://gcc.gnu.org/svn/gcc/branches/gcc-4_7-branch revision 195313 You'll find: gcc-4.7-20130119.tar.bz2 Complete GCC MD5=a9e07a51292d8b84e2b162bf328d06e3 SHA1=a37817fd386afd099d62d026bbc38e9aca695ef3 Diffs from 4.7-20130112 are available in the diffs/ subdirectory. When a particular snapshot is ready for public consumption the LATEST-4.7 link is updated and a message is sent to the gcc list. Please do not use a snapshot before it has been announced that way.
Re: GCC cannot move address calculation to store+load?
A related discussion in this thread: http://gcc.gnu.org/ml/gcc/2013-01/msg00215.html On Fri, Jan 18, 2013 at 7:57 AM, Miles Bader wrote: > Richard Biener writes: >> Note that combine does not apply because %eax is used multiple >> times. This also means that for code-size the combining is not a good >> idea. > > Though the lea instruction seems rather large, so in fact the code is > a fair bit smaller without it, e.g. as generated by clang/llvm: > > clang/llvm 3.1 (-O2 -m32): > > : >0: 8b 44 24 08 mov0x8(%esp),%eax >4: 8b 4c 24 04 mov0x4(%esp),%ecx >8: 8b 14 c1mov(%ecx,%eax,8),%edx >b: 89 15 00 00 00 00 mov%edx,0x0 > 11: 89 04 c1mov%eax,(%ecx,%eax,8) > 14: 31 c0 xor%eax,%eax > 16: c3 ret > > gcc 4.8 20130113 (-O2 -m32): > > : >0: 8b 54 24 08 mov0x8(%esp),%edx >4: 8d 04 d5 00 00 00 00lea0x0(,%edx,8),%eax >b: 03 44 24 04 add0x4(%esp),%eax >f: 8b 08 mov(%eax),%ecx > 11: 89 0d 00 00 00 00 mov%ecx,0x0 > 17: 89 10 mov%edx,(%eax) > 19: 31 c0 xor%eax,%eax > 1b: c3 ret > > -miles > > -- > Ich bin ein Virus. Mach' mit und kopiere mich in Deine .signature.
