http://llvm.org/bugs/show_bug.cgi?id=10552
Duncan Sands <[email protected]> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #3 from Duncan Sands <[email protected]> 2012-11-25 02:53:43 CST --- This was fixed at some point. What now happens at -O0 is that 6 bytes are copied out of the alloca using memcpy into a larger 8 byte alloca, and the larger alloca is used for returning. (You may object that the original alloca only holds 5 bytes of memory, but because it has an alignment of 2 it actually holds 5-bytes-rounded-up-to-alignment = 6 bytes). The optimizers happily turn this into decent code at -O1 and better. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. _______________________________________________ LLVMbugs mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs
