aboutsummaryrefslogtreecommitdiffstats
path: root/kernel/events/hw_breakpoint.c
diff options
context:
space:
mode:
authorOleg Nesterov <oleg@redhat.com>2013-06-20 11:50:11 -0400
committerIngo Molnar <mingo@kernel.org>2013-06-20 11:58:55 -0400
commite1ebe86203e6532eb5a0ae8f26ccae47aca548ae (patch)
treedc9464fa0779f4d45d79f360192a8cf533bb90aa /kernel/events/hw_breakpoint.c
parentf070a4dba984975f6353c6f1d966da1a6ce4b86f (diff)
hw_breakpoint: Simplify list/idx mess in toggle_bp_slot() paths
The enable/disable logic in toggle_bp_slot() is not symmetrical and imho very confusing. "old_count" in toggle_bp_task_slot() is actually new_count because this bp was already removed from the list. Change toggle_bp_slot() to always call list_add/list_del after toggle_bp_task_slot(). This way old_idx is task_bp_pinned() and this entry should be decremented, new_idx is +/-weight and we need to increment this element. The code/logic looks obvious. Reported-by: Vince Weaver <vincent.weaver@maine.edu> Signed-off-by: Oleg Nesterov <oleg@redhat.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Link: http://lkml.kernel.org/r/20130620155011.GA6330@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
Diffstat (limited to 'kernel/events/hw_breakpoint.c')
-rw-r--r--kernel/events/hw_breakpoint.c40
1 files changed, 16 insertions, 24 deletions
diff --git a/kernel/events/hw_breakpoint.c b/kernel/events/hw_breakpoint.c
index ef8ebe560949..dee0148dcf54 100644
--- a/kernel/events/hw_breakpoint.c
+++ b/kernel/events/hw_breakpoint.c
@@ -185,26 +185,20 @@ fetch_this_slot(struct bp_busy_slots *slots, int weight)
185static void toggle_bp_task_slot(struct perf_event *bp, int cpu, bool enable, 185static void toggle_bp_task_slot(struct perf_event *bp, int cpu, bool enable,
186 enum bp_type_idx type, int weight) 186 enum bp_type_idx type, int weight)
187{ 187{
188 unsigned int *tsk_pinned; 188 /* tsk_pinned[n-1] is the number of tasks having n>0 breakpoints */
189 int old_count = 0; 189 unsigned int *tsk_pinned = per_cpu(nr_task_bp_pinned[type], cpu);
190 int old_idx = 0; 190 int old_idx, new_idx;
191 int idx = 0; 191
192 192 old_idx = task_bp_pinned(cpu, bp, type) - 1;
193 old_count = task_bp_pinned(cpu, bp, type); 193 if (enable)
194 old_idx = old_count - 1; 194 new_idx = old_idx + weight;
195 idx = old_idx + weight; 195 else
196 196 new_idx = old_idx - weight;
197 /* tsk_pinned[n] is the number of tasks having n breakpoints */ 197
198 tsk_pinned = per_cpu(nr_task_bp_pinned[type], cpu); 198 if (old_idx >= 0)
199 if (enable) { 199 tsk_pinned[old_idx]--;
200 tsk_pinned[idx]++; 200 if (new_idx >= 0)
201 if (old_count > 0) 201 tsk_pinned[new_idx]++;
202 tsk_pinned[old_idx]--;
203 } else {
204 tsk_pinned[idx]--;
205 if (old_count > 0)
206 tsk_pinned[old_idx]++;
207 }
208} 202}
209 203
210/* 204/*
@@ -228,10 +222,6 @@ toggle_bp_slot(struct perf_event *bp, bool enable, enum bp_type_idx type,
228 } 222 }
229 223
230 /* Pinned counter task profiling */ 224 /* Pinned counter task profiling */
231
232 if (!enable)
233 list_del(&bp->hw.bp_list);
234
235 if (cpu >= 0) { 225 if (cpu >= 0) {
236 toggle_bp_task_slot(bp, cpu, enable, type, weight); 226 toggle_bp_task_slot(bp, cpu, enable, type, weight);
237 } else { 227 } else {
@@ -241,6 +231,8 @@ toggle_bp_slot(struct perf_event *bp, bool enable, enum bp_type_idx type,
241 231
242 if (enable) 232 if (enable)
243 list_add_tail(&bp->hw.bp_list, &bp_task_head); 233 list_add_tail(&bp->hw.bp_list, &bp_task_head);
234 else
235 list_del(&bp->hw.bp_list);
244} 236}
245 237
246/* 238/*