[Bug tree-optimization/81573] [8 Regression] wrong code at -O3 on x86_64-linux-gnu

seurer at gcc dot gnu.org gcc-bugzilla@gcc.gnu.org
Thu Jul 27 14:39:00 GMT 2017


https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81573

--- Comment #7 from seurer at gcc dot gnu.org ---
The problem in the test case is that the code assumes "char" is signed.  On
power it is not.  So "b" starts at 0, is decremented to 255, and ...etc...  
The for loop will run forever.


More information about the Gcc-bugs mailing list