Prev: perf/perf_events: misleading number of samples due to mmap()
Next: [PATCH] x86, hpet.c: Changed type of field work of hpet_work_struct (delayed_work to work_struct)
From: Andi Kleen on 16 Jun 2010 11:00 On Wed, Jun 16, 2010 at 03:48:18PM +0400, Pavel V. Panteleev wrote: > Hello! My name is Pavel V. Panteleev. > > While working with percpu on linux-2.6.35-rc3 I have found trivial bugs in pcpu_build_alloc_info() function: > - we should memset group_cnt to 0 by size of group_cnt, not size of group_map; > - we can delete useless variable group_cnt_max. Thanks for the patch. Tejun maintains percpu code these days, putting him into cc. -Andi Pavel's patch: --- percpu_old.c 2010-06-12 06:14:04.000000000 +0400 +++ percpu_new.c 2010-06-16 15:05:16.280214868 +0400 @@ -1086,7 +1086,7 @@ static int group_map[NR_CPUS] __initdata; static int group_cnt[NR_CPUS] __initdata; const size_t static_size = __per_cpu_end - __per_cpu_start; - int group_cnt_max = 0, nr_groups = 1, nr_units = 0; + int nr_groups = 1, nr_units = 0; size_t size_sum, min_unit_size, alloc_size; int upa, max_upa, uninitialized_var(best_upa); /* units_per_alloc */ int last_allocs, group, unit; @@ -1096,7 +1096,7 @@ /* this function may be called multiple times */ memset(group_map, 0, sizeof(group_map)); - memset(group_cnt, 0, sizeof(group_map)); + memset(group_cnt, 0, sizeof(group_cnt)); /* * Determine min_unit_size, alloc_size and max_upa such that @@ -1130,7 +1130,6 @@ } group_map[cpu] = group; group_cnt[group]++; - group_cnt_max = max(group_cnt_max, group_cnt[group]); } /* -- ak(a)linux.intel.com -- Speaking for myself only. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/ |