Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932205Ab2HQQQU (ORCPT ); Fri, 17 Aug 2012 12:16:20 -0400 Received: from mail-we0-f174.google.com ([74.125.82.174]:49034 "EHLO mail-we0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758343Ab2HQQQD (ORCPT ); Fri, 17 Aug 2012 12:16:03 -0400 Date: Fri, 17 Aug 2012 18:15:57 +0200 From: Frederic Weisbecker To: Michael Neuling Cc: K Prasad , linux-kernel@vger.kernel.org, linuxppc-dev@ozlabs.org, Peter Zijlstra , Ingo Molnar Subject: Re: powerpc/perf: hw breakpoints return ENOSPC Message-ID: <20120817161554.GA3772@somewhere> References: <28857.1345091034@neuling.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <28857.1345091034@neuling.org> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3713 Lines: 99 On Thu, Aug 16, 2012 at 02:23:54PM +1000, Michael Neuling wrote: > Hi, > > I've been trying to get hardware breakpoints with perf to work on POWER7 > but I'm getting the following: > > % perf record -e mem:0x10000000 true > > Error: sys_perf_event_open() syscall returned with 28 (No space left on device). /bin/dmesg may provide additional information. > > Fatal: No CONFIG_PERF_EVENTS=y kernel support configured? > > true: Terminated > > (FWIW adding -a and it works fine) > > Debugging it seems that __reserve_bp_slot() is returning ENOSPC because > it thinks there are no free breakpoint slots on this CPU. > > I have a 2 CPUs, so perf userspace is doing two perf_event_open syscalls > to add a counter to each CPU [1]. The first syscall succeeds but the > second is failing. > > On this second syscall, fetch_bp_busy_slots() sets slots.pinned to be 1, > despite there being no breakpoint on this CPU. This is because the call > the task_bp_pinned, checks all CPUs, rather than just the current CPU. > POWER7 only has one hardware breakpoint per CPU (ie. HBP_NUM=1), so we > return ENOSPC. > > The following patch fixes this by checking the associated CPU for each > breakpoint in task_bp_pinned. I'm not familiar with this code, so it's > provided as a reference to the above issue. > > Mikey > > 1. not sure why it doesn't just do one syscall and specify all CPUs, but > that's another issue. Using two syscalls should work. This patch seems to make sense. I'll try it and run some tests. Can I have your Signed-off-by ? Thanks. > > diff --git a/kernel/events/hw_breakpoint.c b/kernel/events/hw_breakpoint.c > index bb38c4d..e092daa 100644 > --- a/kernel/events/hw_breakpoint.c > +++ b/kernel/events/hw_breakpoint.c > @@ -111,14 +111,16 @@ static unsigned int max_task_bp_pinned(int cpu, enum bp_type_idx type) > * Count the number of breakpoints of the same type and same task. > * The given event must be not on the list. > */ > -static int task_bp_pinned(struct perf_event *bp, enum bp_type_idx type) > +static int task_bp_pinned(int cpu, struct perf_event *bp, enum bp_type_idx type) > { > struct task_struct *tsk = bp->hw.bp_target; > struct perf_event *iter; > int count = 0; > > list_for_each_entry(iter, &bp_task_head, hw.bp_list) { > - if (iter->hw.bp_target == tsk && find_slot_idx(iter) == type) > + if (iter->hw.bp_target == tsk && > + find_slot_idx(iter) == type && > + cpu == iter->cpu) > count += hw_breakpoint_weight(iter); > } > > @@ -141,7 +143,7 @@ fetch_bp_busy_slots(struct bp_busy_slots *slots, struct perf_event *bp, > if (!tsk) > slots->pinned += max_task_bp_pinned(cpu, type); > else > - slots->pinned += task_bp_pinned(bp, type); > + slots->pinned += task_bp_pinned(cpu, bp, type); > slots->flexible = per_cpu(nr_bp_flexible[type], cpu); > > return; > @@ -154,7 +156,7 @@ fetch_bp_busy_slots(struct bp_busy_slots *slots, struct perf_event *bp, > if (!tsk) > nr += max_task_bp_pinned(cpu, type); > else > - nr += task_bp_pinned(bp, type); > + nr += task_bp_pinned(cpu, bp, type); > > if (nr > slots->pinned) > slots->pinned = nr; > @@ -188,7 +190,7 @@ static void toggle_bp_task_slot(struct perf_event *bp, int cpu, bool enable, > int old_idx = 0; > int idx = 0; > > - old_count = task_bp_pinned(bp, type); > + old_count = task_bp_pinned(cpu, bp, type); > old_idx = old_count - 1; > idx = old_idx + weight; > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/