llvm-project/clang/test/CodeGen/asm-2.c

// RUN: %clang_cc1 -emit-llvm %s -o %t -triple i386-pc-linux-gnu -O2
// RUN: not grep "load" %t

// <rdar://problem/6841383>
int cpuid(unsigned data) {
  int a, b;
  
  asm("xyz" :"=a"(a), "=d"(b) : "a"(data));
  return a + b;
}
Update tests to use %clang_cc1 instead of 'clang-cc' or 'clang -cc1'. - This is designed to make it obvious that %clang_cc1 is a "test variable" which is substituted. It is '%clang_cc1' instead of '%clang -cc1' because it can be useful to redefine what gets run as 'clang -cc1' (for example, to set a default target). llvm-svn: 91446 2009-12-16 04:14:24 +08:00			`// RUN: %clang_cc1 -emit-llvm %s -o %t -triple i386-pc-linux-gnu -O2`
Don't use indirect memory destinations for inline asm. Fixes 6841383. llvm-svn: 70523 2009-05-01 08:16:04 +08:00			`// RUN: not grep "load" %t`

			`// <rdar://problem/6841383>`
			`int cpuid(unsigned data) {`
			`int a, b;`

			`asm("xyz" :"=a"(a), "=d"(b) : "a"(data));`
			`return a + b;`
Remove the -arch option from clang-cc: for all practical purposes, it's redundant with -triple. llvm-svn: 72108 2009-05-19 19:12:40 +08:00			`}`