Fix the build with Clang 3.9

Clang 3.9 complains that `fild` is ambiguous with respect to the size of its
memory operand.  Specify `fildq` explicitly.

Drop "-no-integrated-as" from the docs, as both Clang 3.5 and 3.9 work fine
with or without it.

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>

diff --git a/docs/mainpage.dox b/docs/mainpage.dox
index a16cda9f1a73a1b6ca2f71d8bc6373447b569701..75294a0996508fb64dd702b5bc87a6ce719cd9d6 100644 (file)
--- a/docs/mainpage.dox
+++ b/docs/mainpage.dox
@@ -35,7 +35,7 @@ Environment | Guest | Width | Paging
 Requirements:
 - GNU Make >= 3.81
 - GNU comaptible toolchain, capable of `-std=gnu99`, `-m64`, and `-m32`
-    - Clang may be used, via `CC="clang -no-integrated-as"`
+    - Clang may be used, via `CC="clang"`
 - Python
 
 To obtain and build:

diff --git a/tests/fpu-exception-emulation/main.c b/tests/fpu-exception-emulation/main.c
index 7a0379c3fd070768e641ed5380b3a77f668e50ff..ea4d4690e196bbf986caded5f7adc87eb49f1908 100644 (file)
--- a/tests/fpu-exception-emulation/main.c
+++ b/tests/fpu-exception-emulation/main.c
@@ -77,7 +77,7 @@ unsigned int probe_x87(bool force)
     asm volatile ("test %[fep], %[fep];"
                   "jz 1f;"
                   _ASM_XEN_FEP
-                  "1: fild %[ptr]; 2:"
+                  "1: fildq %[ptr]; 2:"
                   _ASM_EXTABLE_HANDLER(1b, 2b, ex_record_fault_eax)
                   : "+a" (fault)
                   : [ptr] "m" (zero),