perf_event: x86: Allocate the fake_cpuc
Peter Zijlstra [Fri, 22 Jan 2010 13:35:46 +0000 (14:35 +0100)]
GCC was complaining the stack usage was too large, so allocate the
structure.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
LKML-Reference: <20100122155535.411197266@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

arch/x86/kernel/cpu/perf_event.c

index 0bd23d0..7bd359a 100644 (file)
@@ -2606,10 +2606,13 @@ static const struct pmu pmu = {
 static int validate_group(struct perf_event *event)
 {
        struct perf_event *leader = event->group_leader;
-       struct cpu_hw_events fake_cpuc;
-       int n;
+       struct cpu_hw_events *fake_cpuc;
+       int ret, n;
 
-       memset(&fake_cpuc, 0, sizeof(fake_cpuc));
+       ret = -ENOMEM;
+       fake_cpuc = kmalloc(sizeof(*fake_cpuc), GFP_KERNEL | __GFP_ZERO);
+       if (!fake_cpuc)
+               goto out;
 
        /*
         * the event is not yet connected with its
@@ -2617,18 +2620,24 @@ static int validate_group(struct perf_event *event)
         * existing siblings, then add the new event
         * before we can simulate the scheduling
         */
-       n = collect_events(&fake_cpuc, leader, true);
+       ret = -ENOSPC;
+       n = collect_events(fake_cpuc, leader, true);
        if (n < 0)
-               return -ENOSPC;
+               goto out_free;
 
-       fake_cpuc.n_events = n;
-       n = collect_events(&fake_cpuc, event, false);
+       fake_cpuc->n_events = n;
+       n = collect_events(fake_cpuc, event, false);
        if (n < 0)
-               return -ENOSPC;
+               goto out_free;
 
-       fake_cpuc.n_events = n;
+       fake_cpuc->n_events = n;
 
-       return x86_schedule_events(&fake_cpuc, n, NULL);
+       ret = x86_schedule_events(fake_cpuc, n, NULL);
+
+out_free:
+       kfree(fake_cpuc);
+out:
+       return ret;
 }
 
 const struct pmu *hw_perf_event_init(struct perf_event *event)