On Fri, Sep 19, 2014 at 01:26:03PM -0400, Pranith Kumar wrote:
> __this_cpu_{read/write}() uses sizeof() to determine the size of the variable.
> Using this operation on a bool type causes sparse to complain.
>
> I checked the generated code and there is only a slight difference in
> instructions generated.
>
> I am not sure this warning by sparse is really valid, but let us silence it
> anyways.
>
> Signed-off-by: Pranith Kumar <bobby.pr...@gmail.com>
I believe this warning was actually fixed with current versions of
Sparse; there was some discussion and a patch on the linux-sparse
mailing list. Can you please recheck with the latest Sparse from git,
and if it still breaks, please raise that on linux-sparse?
> kernel/rcu/tree.c | 12 ++++++++----
> kernel/rcu/tree_plugin.h | 6 ++++--
> 2 files changed, 12 insertions(+), 6 deletions(-)
>
> diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c
> index bc085e5..85f228a 100644
> --- a/kernel/rcu/tree.c
> +++ b/kernel/rcu/tree.c
> @@ -186,21 +186,25 @@ static int rcu_gp_in_progress(struct rcu_state *rsp)
> */
> void rcu_sched_qs(void)
> {
> - if (!__this_cpu_read(rcu_sched_data.passed_quiesce)) {
> + struct rcu_data *rdp = __this_cpu_ptr(&rcu_sched_data);
> +
> + if (!rdp->passed_quiesce) {
> trace_rcu_grace_period(TPS("rcu_sched"),
> __this_cpu_read(rcu_sched_data.gpnum),
> TPS("cpuqs"));
> - __this_cpu_write(rcu_sched_data.passed_quiesce, 1);
> + rdp->passed_quiesce = true;
> }
> }
>
> void rcu_bh_qs(void)
> {
> - if (!__this_cpu_read(rcu_bh_data.passed_quiesce)) {
> + struct rcu_data *rdp = __this_cpu_ptr(&rcu_bh_data);
> +
> + if (!rdp->passed_quiesce) {
> trace_rcu_grace_period(TPS("rcu_bh"),
> __this_cpu_read(rcu_bh_data.gpnum),
> TPS("cpuqs"));
> - __this_cpu_write(rcu_bh_data.passed_quiesce, 1);
> + rdp->passed_quiesce = true;
> }
> }
>
> diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h
> index 77f9376..98de303 100644
> --- a/kernel/rcu/tree_plugin.h
> +++ b/kernel/rcu/tree_plugin.h
> @@ -143,11 +143,13 @@ EXPORT_SYMBOL_GPL(rcu_batches_completed);
> */
> static void rcu_preempt_qs(void)
> {
> - if (!__this_cpu_read(rcu_preempt_data.passed_quiesce)) {
> + struct rcu_data *rdp = __this_cpu_ptr(&rcu_preempt_data);
> +
> + if (!rdp->passed_quiesce) {
> trace_rcu_grace_period(TPS("rcu_preempt"),
> __this_cpu_read(rcu_preempt_data.gpnum),
> TPS("cpuqs"));
> - __this_cpu_write(rcu_preempt_data.passed_quiesce, 1);
> + rdp->passed_quiesce = true;
> barrier(); /* Coordinate with rcu_preempt_check_callbacks(). */
> current->rcu_read_unlock_special.b.need_qs = false;
> }
> --
> 2.1.0
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
