http://llvm.org/bugs/show_bug.cgi?id=18916
Reid Kleckner <[email protected]> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |RESOLVED Resolution|--- |FIXED --- Comment #7 from Reid Kleckner <[email protected]> --- (In reply to comment #5) > $ cat t.cpp > int main() { > asm (".intel_syntax noprefix;"); > asm ("movdqu xmm4, [rcx];"); > asm (".att_syntax prefix;"); > } > > $ g++ t.cpp > > $ clang++ t.cpp > t.cpp:3:8: error: invalid operand for instruction > asm ("movdqu xmm4, [rcx];"); > ^ > <inline asm>:1:15: note: instantiated into assembly here > movdqu xmm4, [rcx]; > ^~~~~~ > 1 error generated. I'm pretty sure we don't support preserving the assembly dialect across asm blob invocations when using the integrated assembler. The second blob is being parsed as AT&T style asm. Merge the blobs to get the desired behavior. -- 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
