x86/headers/UAPI: Use __u64 instead of u64 in <uapi/asm/hyperv.h>

... since u64 has a hidden header dependency that was not there before
using it (i.e. it breaks our VMM build).

Also, __u64 is the right way to expose data types through UAPI.

Signed-off-by: KarimAllah Ahmed <[email protected]>
Acked-by: Thomas Gleixner <[email protected]>
Cc: Haiyang Zhang <[email protected]>
Cc: K. Y. Srinivasan <[email protected]>
Cc: Linus Torvalds <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Cc: Stephen Hemminger <[email protected]>
Cc: [email protected]
Fixes: 9328626 ("x86/hyperv: Reenlightenment notifications support")
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Ingo Molnar <[email protected]>

Author: KarimAllah Ahmed

arch/x86/include/uapi/asm/hyperv.h

@@ -241,24 +241,24 @@
 #define HV_X64_MSR_REENLIGHTENMENT_CONTROL	0x40000106
 
 struct hv_reenlightenment_control {
-	u64 vector:8;
-	u64 reserved1:8;
-	u64 enabled:1;
-	u64 reserved2:15;
-	u64 target_vp:32;
+	__u64 vector:8;
+	__u64 reserved1:8;
+	__u64 enabled:1;
+	__u64 reserved2:15;
+	__u64 target_vp:32;
 };
 
 #define HV_X64_MSR_TSC_EMULATION_CONTROL	0x40000107
 #define HV_X64_MSR_TSC_EMULATION_STATUS		0x40000108
 
 struct hv_tsc_emulation_control {
-	u64 enabled:1;
-	u64 reserved:63;
+	__u64 enabled:1;
+	__u64 reserved:63;
 };
 
 struct hv_tsc_emulation_status {
-	u64 inprogress:1;
-	u64 reserved:63;
+	__u64 inprogress:1;
+	__u64 reserved:63;
 };
 
 #define HV_X64_MSR_HYPERCALL_ENABLE		0x00000001