On Wed, May 14, 2014 at 05:40:50AM +0200, Greg KH wrote: > Peter, Guenter is reporting that the patch below breaks all non-x86 > builds with: > > include/asm-generic/preempt.h: In function '__preempt_count_dec_and_test': > include/asm-generic/preempt.h:71:2: error: implicit declaration of function > 'tif_need_resched' [-Werror=implicit-function-declaration] > return !--*preempt_count_ptr() && tif_need_resched(); > > Any ideas?
Yeah, I suck :/ Ok, the below seems to build on a number of !x86 archs and still does build for i386 too. Now its a little fragile, but since there's only a single <asm/preempt.h> include (from linux/preempt.h) it should be safe. Obviously its not actually been tested (as per the tags) because the patch is different. --- Subject: x86,preempt: Fix preemption for i386 From: Peter Zijlstra <pet...@infradead.org> Date: Wed, 9 Apr 2014 16:24:47 +0200 Many people reported preemption/reschedule problems with i386 kernels for .13 and .14. After Michele bisected this to a combination of 3e8e42c69bb ("sched: Revert need_resched() to look at TIF_NEED_RESCHED") ded79754754 ("irq: Force hardirq exit's softirq processing on its own stack") it finally dawned on me that i386's current_thread_info() was to blame. When we are on interrupt/exception stacks, we fail to observe the right TIF_NEED_RESCHED bit and therefore the PREEMPT_NEED_RESCHED folding malfunctions. Current upstream fixes this by making i386 behave the same as x86_64 already did: 2432e1364bbe ("x86: Nuke the supervisor_stack field in i386 thread_info") b807902a88c4 ("x86: Nuke GET_THREAD_INFO_WITH_ESP() macro for i386") 0788aa6a23cb ("x86: Prepare removal of previous_esp from i386 thread_info structure") 198d208df437 ("x86: Keep thread_info on thread stack in x86_32") However, that is far too much to stuff into -stable. Therefore I propose we merge the below patch which uses task_thread_info(current) for tif_need_resched() instead of the ESP based current_thread_info(). This makes sure we always observe the one true TIF_NEED_RESCHED bit and things will work as expected again. Cc: Greg KH <g...@kroah.com> Cc: b...@alien8.de Cc: fweis...@gmail.com Cc: david.a.co...@linux.intel.com Cc: mi...@kernel.org Cc: fweis...@gmail.com Cc: g...@kroah.com Cc: Steven Rostedt <rost...@goodmis.org> Cc: gre...@linuxfoundation.org Cc: pbonz...@redhat.com Cc: rost...@goodmis.org Cc: stefan.ba...@canonical.com Cc: mi...@kernel.org Cc: toralf.foers...@gmx.de Cc: David Cohen <david.a.co...@linux.intel.com> Cc: Steven Rostedt <rost...@goodmis.org> Cc: torva...@linux-foundation.org Cc: Paolo Bonzini <pbonz...@redhat.com> Cc: David Cohen <david.a.co...@linux.intel.com> Cc: <stable@vger.kernel.org> Cc: Borislav Petkov <b...@alien8.de> Cc: Paolo Bonzini <pbonz...@redhat.com> Cc: <stable-comm...@vger.kernel.org> Cc: Borislav Petkov <b...@alien8.de> Cc: pet...@infradead.org Cc: Linus Torvalds <torva...@linux-foundation.org> Cc: barra_c...@katamail.com Tested-by: Stefan Bader <stefan.ba...@canonical.com> Tested-by: Toralf F�rster <toralf.foers...@gmx.de> Tested-by: Michele Ballabio <barra_c...@katamail.com> Signed-off-by: Greg Kroah-Hartman <gre...@linuxfoundation.org> Signed-off-by: Peter Zijlstra <pet...@infradead.org> Link: http://lkml.kernel.org/r/20140409142447.gd13...@twins.programming.kicks-ass.net --- arch/x86/include/asm/preempt.h | 12 ++++++++++++ include/linux/preempt.h | 2 ++ include/linux/thread_info.h | 2 -- 3 files changed, 14 insertions(+), 2 deletions(-) --- a/arch/x86/include/asm/preempt.h +++ b/arch/x86/include/asm/preempt.h @@ -5,6 +5,18 @@ #include <asm/percpu.h> #include <linux/thread_info.h> +#ifdef CONFIG_X86_32 +/* + * i386's current_thread_info() depends on ESP and for interrupt/exception + * stacks this doesn't yield the actual task thread_info. + * + * We hard rely on the fact that all the TIF_NEED_RESCHED bits are + * the same, therefore use the slightly more expensive version below. + */ +#undef tif_need_resched +#define tif_need_resched() test_tsk_thread_flag(current, TIF_NEED_RESCHED) +#endif + DECLARE_PER_CPU(int, __preempt_count); /* --- a/include/linux/preempt.h +++ b/include/linux/preempt.h @@ -15,6 +15,8 @@ */ #define PREEMPT_NEED_RESCHED 0x80000000 +#define tif_need_resched() test_thread_flag(TIF_NEED_RESCHED) + #include <asm/preempt.h> #if defined(CONFIG_DEBUG_PREEMPT) || defined(CONFIG_PREEMPT_TRACER) --- a/include/linux/thread_info.h +++ b/include/linux/thread_info.h @@ -118,8 +118,6 @@ static inline __deprecated void set_need */ } -#define tif_need_resched() test_thread_flag(TIF_NEED_RESCHED) - #if defined TIF_RESTORE_SIGMASK && !defined HAVE_SET_RESTORE_SIGMASK /* * An arch can define its own version of set_restore_sigmask() to get the
pgpCAbRtFAaf3.pgp
Description: PGP signature