[NFC][Clang] Modify expect of fail test or XFAIL because CSKY align is different

[zixuan-wu]

Summary

CSKY is always in 4-byte align, no matter it's long long type. For global aggregate variable, it's 4-byte align if its size is bigger than or equal to 4 bytes.

[rengolin]

clang/test/Sema/builtin-alloca-with-align.c

32

This test is platform agnostic, perhaps the extra error could be in a new test, exclusively for csky?

[zixuan-wu]

clang/test/Sema/builtin-alloca-with-align.c

32

Then I prefer to split the file and add UNSUPPORTED for CSKY in the other file which only contains test8

[rengolin]

clang/test/Sema/builtin-alloca-with-align.c

32

But then you wouldn't be testing the extra error you want... hmm.

Maybe it would be fine the way you did it originally.

Would be nice to get a clang person to give their opinion.

[zixuan-wu]

Could anybody else have a review or nominate a reviewer?

[aaron.ballman]

clang/test/Sema/builtin-alloca-with-align.c

32

// Comment which explains why this is special.

I think I would test this slightly differently:

void test8(void) {
  __builtin_alloca_with_align(sizeof(__INT64_TYPE__), __alignof__(__INT64_TYPE__)); // expected-warning {{second argument to __builtin_alloca_with_align is supposed to be in bits}}
#ifdef __csky__
  // expected-error@-2 {{requested alignment must be 8 or greater}}
  // Comment which explains why this is special.
#endif // __csky__
}

[zixuan-wu]

clang/test/Sema/builtin-alloca-with-align.c

32

Reusing the first warning is good.

[zixuan-wu]

Address comments.

[zixuan-wu]

Gentle pin..

[aaron.ballman]

LGTM!

clang/test/CodeGen/c-strings.c

 // MSABI: @f4.x = internal global %struct.s { i8* getelementptr inbounds ([6 x i8], [6 x i8]* @"??_C@_05CJBACGMB@hello?$AA@", i32 0, i32 0) }
 // CHECK: @x = {{(dso_local )?}}global [3 x i8] c"ola", align [[ALIGN]]
 
 // XFAIL: hexagon
 // Hexagon aligns arrays of size 8+ bytes to a 64-bit boundary, which
 // fails the check for "@f3.x = ... align [ALIGN]", since ALIGN is derived
 // from the alignment of a single i8, which is still 1.
 
+// XFAIL: csky
+// CSKY aligns arrays of size 4+ bytes to a 32-bit boundary, which
+// fails the check for "@f2.x = ... align [ALIGN]", since ALIGN is derived
+// from the alignment of a single i8, which is still 1.
+
 #if defined(__s390x__)
 unsigned char align = 2;
 #else
 unsigned char align = 1;
 #endif
 
 void bar(const char *);
 
void f4(void) {
   static struct s {
     char *name;
   } x = { "hello" };
   gaz(&x);
   // CHECK: call {{.*}}void @gaz({{.*}} @f4.x
 }
 
 char x[3] = "ola";
-

clang/test/Sema/builtin-alloca-with-align.c

void test6(int a, int j) {
   __builtin_alloca_with_align(a, j); // expected-error {{must be a constant integer}}
 }
 
 void test7(int a) {
   __builtin_alloca_with_align(a, 2); // expected-error {{requested alignment must be 8 or greater}}
 }
 
 void test8(void) {
   __builtin_alloca_with_align(sizeof(__INT64_TYPE__), __alignof__(__INT64_TYPE__)); // expected-warning {{second argument to __builtin_alloca_with_align is supposed to be in bits}}
+#if defined(__csky__)
+  // expected-error@-1 {{requested alignment must be 8 or greater}}
+  // Because the alignment of long long is 4 in CSKY target
+#endif
 }