c/2721: Linux kernel 2.4.4 compile failure

[dmuell@gmx.net]

>Number:         2721
>Category:       c
>Synopsis:       compile failure of Linux 2.4.4 kernel
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          rejects-legal
>Submitter-Id:   net
>Arrival-Date:   Thu May 03 01:26:00 PDT 2001
>Closed-Date:
>Last-Modified:
>Originator:     Dirk Mueller
>Release:        3.0 20010502 (prerelease)
>Organization:
>Environment:
System: Linux rotes20 2.4.4 #4 Mon Apr 30 21:55:44 CEST 2001 i686 unknown
Architecture: i686

	
host: i686-pc-linux-gnu
build: i686-pc-linux-gnu
target: i686-pc-linux-gnu
configured with: ../configure --enable-shared --enable-threads --enable-languages=c,c++
>Description:

with i386 / Athlon / Duron switch enabled in linux configuration, you get this
error while compiling: 

ake[2]: Entering directory `/usr/src/linux/kernel'
gcc -D__KERNEL__ -I/usr/src/linux/include -Wall -Wstrict-prototypes -O2
-fomit-frame-pointer -fno-strict-aliasing -pipe -mpreferred-stack-boundary=2
-march=athlon     -DEXPORT_SYMTAB -c sys.c
sys.c: In function `sys_gethostname':
/usr/src/linux/include/asm/rwsem.h:152: inconsistent operand constraints in an
`asm'

The error can be worked around by replacing first "+d" constrain by a "=d" and
adding a "0" constrain as input. 

However, this can't be the right solution as there are other macros in the same
file that use _exactly_ the same constraints and only differ by the
initialisation of the long count and those are accepted. So I suspect it to be
a bug in gcc instead. 

>How-To-Repeat:
>Fix:
>Release-Note:
>Audit-Trail:
>Unformatted: