On Mo, 2010-10-18 at 20:54 +0200, Peter Zijlstra wrote: > On Fri, 2010-10-15 at 13:07 +0200, Gerald Schaefer wrote: > > From: Gerald Schaefer <gerald.schaefer@xxxxxxxxxx> > > > > The spinning mutex implementation uses cpu_relax() in busy loops as a > > compiler barrier. Depending on the architecture, cpu_relax() may do more > > than needed in this specific mutex spin loops. On System z we also give > > up the time slice of the virtual cpu in cpu_relax(), which prevents > > effective spinning on the mutex. > > > > This patch replaces cpu_relax() in the spinning mutex code with > > arch_mutex_cpu_relax(), which can be defined by each architecture in > > include/asm/mutex.h. The default is still cpu_relax(), so this should > > not affect other architectures than System z for now. > > Ingo's randconfig build found the following, .config attached. > > including "asm/mutex.h" isn't advised. > > CC kernel/mutex.o > In file included from /usr/src/linux-2.6/kernel/mutex.c:33: > /usr/src/linux-2.6/include/asm-generic/mutex-null.h:13:1: warning: "__mutex_fastpath_lock" redefined > In file included from /usr/src/linux-2.6/arch/x86/include/asm/mutex.h:4, > from /usr/src/linux-2.6/include/linux/mutex.h:19, > from /usr/src/linux-2.6/kernel/mutex.c:20: > /usr/src/linux-2.6/arch/x86/include/asm/mutex_64.h:19:1: warning: this is the location of the previous definition > In file included from /usr/src/linux-2.6/kernel/mutex.c:33: > /usr/src/linux-2.6/include/asm-generic/mutex-null.h:15:1: warning: "__mutex_fastpath_unlock" redefined > In file included from /usr/src/linux-2.6/arch/x86/include/asm/mutex.h:4, > from /usr/src/linux-2.6/include/linux/mutex.h:19, > from /usr/src/linux-2.6/kernel/mutex.c:20: > /usr/src/linux-2.6/arch/x86/include/asm/mutex_64.h:62:1: warning: this is the location of the previous definition > In file included from /usr/src/linux-2.6/kernel/mutex.c:33: > /usr/src/linux-2.6/include/asm-generic/mutex-null.h:13:1: warning: "__mutex_fastpath_lock" redefined > In file included from /usr/src/linux-2.6/arch/x86/include/asm/mutex.h:4, > from /usr/src/linux-2.6/include/linux/mutex.h:19, > from /usr/src/linux-2.6/kernel/mutex.c:20: > /usr/src/linux-2.6/arch/x86/include/asm/mutex_64.h:19:1: warning: this is the location of the previous definition > In file included from /usr/src/linux-2.6/kernel/mutex.c:33: > /usr/src/linux-2.6/include/asm-generic/mutex-null.h:15:1: warning: "__mutex_fastpath_unlock" redefined > In file included from /usr/src/linux-2.6/arch/x86/include/asm/mutex.h:4, > from /usr/src/linux-2.6/include/linux/mutex.h:19, > from /usr/src/linux-2.6/kernel/mutex.c:20: > /usr/src/linux-2.6/arch/x86/include/asm/mutex_64.h:62:1: warning: this is the location of the previous definition Ok, I see now that including <asm/mutex.h> from include/linux/mutex.h is not a good idea, because of this code in kernel/mutex.c (and the conflict with CONFIG_DEBUG_MUTEXES set): #ifdef CONFIG_DEBUG_MUTEXES # include "mutex-debug.h" # include <asm-generic/mutex-null.h> #else # include "mutex.h" # include <asm/mutex.h> #endif Putting the architecture specific details of arch_mutex_cpu_relax() somewhere else than <asm/mutex.h> doesn't seem like a good idea either. Also, putting an "#ifndef CONFIG_DEBUG_MUTEXES" around my "#include <asm/mutex.h>" in include/linux/mutex.h would fix the conflict, but that also looks rather ugly. So I guess I'll just go back to the original Kconfig approach, which at least avoids all this header file mess. I'll send a new patch. -- To unsubscribe from this list: send the line "unsubscribe linux-s390" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
