|
From: Patrick J. L. <lop...@gm...> - 2012-02-16 02:26:13
|
Hello. I downloaded the development version of valgrind from Subversion today. It configured and built fine, but while trying to run it on my application, I got this error during the run: vex amd64->IR: unhandled instruction bytes: 0xC2 0x0 0x0 0x66 0x66 0x66 0x90 0x66 ==10739== valgrind: Unrecognised instruction at address 0x1ff07b5. ==10739== at 0x1FF07B5: __intel_get_new_mem_ops_cpuid (in myApp) Disassembly of myApp leading up to 0x1FF07B5: 0x1ff07aa <__intel_get_new_mem_ops_cpuid+26>: mov %edx,0xc(%r9) 0x1ff07ae <__intel_get_new_mem_ops_cpuid+30>: pop %rdx 0x1ff07af <__intel_get_new_mem_ops_cpuid+31>: pop %rcx 0x1ff07b0 <__intel_get_new_mem_ops_cpuid+32>: pop %rbx 0x1ff07b1 <__intel_get_new_mem_ops_cpuid+33>: pop %rax 0x1ff07b2 <__intel_get_new_mem_ops_cpuid+34>: mov %r8d,%eax 0x1ff07b5 <__intel_get_new_mem_ops_cpuid+37>: retq $0x0 Is this the correct forum for reporting bugs in the development version? Thanks! - Pat |
|
From: John R. <jr...@bi...> - 2012-02-16 05:50:10
|
> Hello. I downloaded the development version of valgrind from > Subversion today. Thank you for mentioning the version of valgrind! > It configured and built fine, but while trying to > run it on my application, I got this error during the run: > > vex amd64->IR: unhandled instruction bytes: 0xC2 0x0 0x0 0x66 0x66 > 0x66 0x90 0x66 > ==10739== valgrind: Unrecognised instruction at address 0x1ff07b5. > ==10739== at 0x1FF07B5: __intel_get_new_mem_ops_cpuid (in myApp) > 0x1ff07b5 <__intel_get_new_mem_ops_cpuid+37>: retq $0x0 Yes, 0xC2 is unhandled. Gcc generates that almost never; which compiler did? Also, "ret $0" is bigger and slower than the equivalent "ret". > Is this the correct forum for reporting bugs in the development version? It's fine to post here to see if there might be a "community" for this bug, particularly if you name compiler and the origin of the source code. But if you really want it fixed, then please report an official bug, else the issue *will* become lost, forgotten, and ignored. Click the "Bug Reports" link in the Contacts sidebar at http://valgrind.org . -- |
|
From: Tom H. <to...@co...> - 2012-02-16 07:16:08
|
On 16/02/12 02:26, Patrick J. LoPresti wrote: > Disassembly of myApp leading up to 0x1FF07B5: > > 0x1ff07aa<__intel_get_new_mem_ops_cpuid+26>: mov %edx,0xc(%r9) > 0x1ff07ae<__intel_get_new_mem_ops_cpuid+30>: pop %rdx > 0x1ff07af<__intel_get_new_mem_ops_cpuid+31>: pop %rcx > 0x1ff07b0<__intel_get_new_mem_ops_cpuid+32>: pop %rbx > 0x1ff07b1<__intel_get_new_mem_ops_cpuid+33>: pop %rax > 0x1ff07b2<__intel_get_new_mem_ops_cpuid+34>: mov %r8d,%eax > 0x1ff07b5<__intel_get_new_mem_ops_cpuid+37>: retq $0x0 > > Is this the correct forum for reporting bugs in the development version? Well the bug tracker is probably better, but it's already done: https://bugs.kde.org/show_bug.cgi?id=292430 Tom -- Tom Hughes (to...@co...) http://compton.nu/ |
|
From: Julian S. <js...@ac...> - 2012-02-16 19:19:37
|
I just committed a fix (vex r2255). J On Thursday, February 16, 2012, Tom Hughes wrote: > On 16/02/12 02:26, Patrick J. LoPresti wrote: > > Disassembly of myApp leading up to 0x1FF07B5: > > 0x1ff07aa<__intel_get_new_mem_ops_cpuid+26>: mov > > %edx,0xc(%r9) 0x1ff07ae<__intel_get_new_mem_ops_cpuid+30>: > > pop %rdx 0x1ff07af<__intel_get_new_mem_ops_cpuid+31>: pop > > %rcx 0x1ff07b0<__intel_get_new_mem_ops_cpuid+32>: pop > > %rbx 0x1ff07b1<__intel_get_new_mem_ops_cpuid+33>: pop %rax > > 0x1ff07b2<__intel_get_new_mem_ops_cpuid+34>: mov %r8d,%eax > > 0x1ff07b5<__intel_get_new_mem_ops_cpuid+37>: retq $0x0 > > > > Is this the correct forum for reporting bugs in the development version? > > Well the bug tracker is probably better, but it's already done: > > https://bugs.kde.org/show_bug.cgi?id=292430 > > Tom |