This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: PATCH: Check __x86_64__ instead of __WORDSIZE in bits/setjmp.h

On Mon, Mar 19, 2012 at 4:07 PM, Roland McGrath <roland@hack.frob.com> wrote:
>> > We need to be very sure this cannot perturb anyone's ABI,
>> > e.g. mangled names and the like.
>> >
>> > I'm not aware of any such issue with __jmp_buf in particular, but
>> > we must be circumspect.
>
> The patch looks OK now as far as nits go. ?But I want to let some other
> people give opinions about the safety vs name mangling et al.
>

How about this patch?  It defines

typedef long long int __jmp_buf[8];

only for x86-64 with __WORDSIZE != 64.

Thanks.

-- 
H.J.

2012-03-19  H.J. Lu  <hongjiu.lu@intel.com>

	* sysdeps/x86_64/bits/setjmp.h (__jmp_buf): Support x86-64 with
	__WORDSIZE != 64.

diff --git a/sysdeps/x86_64/bits/setjmp.h b/sysdeps/x86_64/bits/setjmp.h
index a66ca0f..bb6920d 100644
--- a/sysdeps/x86_64/bits/setjmp.h
+++ b/sysdeps/x86_64/bits/setjmp.h
@@ -29,6 +29,8 @@
 
 # if __WORDSIZE == 64
 typedef long int __jmp_buf[8];
+# elif defined  __x86_64__
+typedef long long int __jmp_buf[8];
 # else
 typedef int __jmp_buf[6];
 # endif
Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]