Bug 288274

Version:           3.5.0
OS:                Linux

I'm using valgrind on code that's linked with the Intel Performance Primitives (IPP).  When running valgrind I get the following error.  If I rewrite the code not to use IPP, the problem does not occur.

vex amd64->IR: unhandled instruction bytes: 0x48 0xF 0x5A 0x7 0x48 0xF
==13887== valgrind: Unrecognised instruction at address 0x1731020.
==13887== Your program just tried to execute an instruction that Valgrind
==13887== did not recognise.  There are two possible reasons for this.
==13887== 1. Your program has a bug and erroneously jumped to a non-code
==13887==    location.  If you are running Memcheck and you just saw a
==13887==    warning about a bad jump, it's probably your program's fault.
==13887== 2. The instruction is legitimate but Valgrind doesn't handle it,
==13887==    i.e. it's Valgrind's fault.  If you think this is the case or
==13887==    you are not sure, please let us know and we'll try to fix it.
==13887== Either way, Valgrind will now raise a SIGILL signal which will
==13887== probably kill your program.
==13887== 
==13887== Process terminating with default action of signal 4 (SIGILL): dumping core
==13887==  Illegal opcode at address 0x1731020
==13887==    at 0x1731020: ??? (in /home/user/project/bin/program)
==13887==    by 0x199670E: u8_ownippsSum_32f (in /home/user/project/bin/program)
==13887==    by 0x191453C: u8_ownippsMean_32f (in /home/user/project/bin/program)
==13887==    by 0x1735E81: u8_ippsMean_32f (in /home/user/project/bin/program)




Reproducible: Always

Steps to Reproduce:
Build code, and run it with valgrind.

Actual Results:  
I get the crash reported above.

Expected Results:  
I expect valgrind to keep executing.