Re: [PATCH -next] riscv: perf_event: Make some funciton static

From: Palmer Dabbelt
Date: Mon May 11 2020 - 17:07:05 EST


On Sun, 10 May 2020 18:30:36 PDT (-0700), wangkefeng.wang@xxxxxxxxxx wrote:

On 2020/5/9 1:13, Palmer Dabbelt wrote:
What's the "-next" for? This seems appropriate for an RC to me, as
it's a
build fix.

Thanks for your review, this patch and "[PATCH -next] riscv: perf:
RISCV_BASE_PMU

should be closeable", I fix the issue based on linux-next, so add the
next prefix ; )

OK, makes sense. I've put it on fixes.


and we also found some another build issue when add RISCV arch to huawei
build robot,

will send out the patches later.

Thanks!


On Thu, 07 May 2020 08:04:44 PDT (-0700), wangkefeng.wang@xxxxxxxxxx
wrote:
Fixes the following warning detected when running make with W=1,
../arch/riscv/kernel/perf_event.c:150:5: warning: no previous
prototype for âriscv_map_cache_decodeâ [-Wmissing-prototypes]
Âint riscv_map_cache_decode(u64 config, unsigned int *type,
ÂÂÂÂ ^~~~~~~~~~~~~~~~~~~~~~
../arch/riscv/kernel/perf_event.c:345:13: warning: no previous
prototype for âriscv_base_pmu_handle_irqâ [-Wmissing-prototypes]
Âirqreturn_t riscv_base_pmu_handle_irq(int irq_num, void *dev)
ÂÂÂÂÂÂÂÂÂÂÂÂ ^~~~~~~~~~~~~~~~~~~~~~~~~
../arch/riscv/kernel/perf_event.c:364:6: warning: no previous
prototype for ârelease_pmc_hardwareâ [-Wmissing-prototypes]
Âvoid release_pmc_hardware(void)
ÂÂÂÂÂ ^~~~~~~~~~~~~~~~~~~~
../arch/riscv/kernel/perf_event.c:467:12: warning: no previous
prototype for âinit_hw_perf_eventsâ [-Wmissing-prototypes]
Âint __init init_hw_perf_events(void)
ÂÂÂÂÂÂÂÂÂÂÂ ^~~~~~~~~~~~~~~~~~~

Cc: Alan Kao <alankao@xxxxxxxxxxxxx>
Signed-off-by: Kefeng Wang <wangkefeng.wang@xxxxxxxxxx>
---
Âarch/riscv/kernel/perf_event.c | 8 ++++----
Â1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/arch/riscv/kernel/perf_event.c
b/arch/riscv/kernel/perf_event.c
index 91626d9ae5f2..c835f0362d94 100644
--- a/arch/riscv/kernel/perf_event.c
+++ b/arch/riscv/kernel/perf_event.c
@@ -147,7 +147,7 @@ static int riscv_map_hw_event(u64 config)
ÂÂÂÂ return riscv_pmu->hw_events[config];
Â}

-int riscv_map_cache_decode(u64 config, unsigned int *type,
+static int riscv_map_cache_decode(u64 config, unsigned int *type,
ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ unsigned int *op, unsigned int *result)
Â{
ÂÂÂÂ return -ENOENT;
@@ -342,7 +342,7 @@ static void riscv_pmu_del(struct perf_event
*event, int flags)

Âstatic DEFINE_MUTEX(pmc_reserve_mutex);

-irqreturn_t riscv_base_pmu_handle_irq(int irq_num, void *dev)
+static irqreturn_t riscv_base_pmu_handle_irq(int irq_num, void *dev)
Â{
ÂÂÂÂ return IRQ_NONE;
Â}
@@ -361,7 +361,7 @@ static int reserve_pmc_hardware(void)
ÂÂÂÂ return err;
Â}

-void release_pmc_hardware(void)
+static void release_pmc_hardware(void)
Â{
ÂÂÂÂ mutex_lock(&pmc_reserve_mutex);
ÂÂÂÂ if (riscv_pmu->irq >= 0)
@@ -464,7 +464,7 @@ static const struct of_device_id
riscv_pmu_of_ids[] = {
ÂÂÂÂ { /* sentinel value */ }
Â};

-int __init init_hw_perf_events(void)
+static int __init init_hw_perf_events(void)
Â{
ÂÂÂÂ struct device_node *node = of_find_node_by_type(NULL, "pmu");
ÂÂÂÂ const struct of_device_id *of_id;

Reviewed-by: Palmer Dabbelt <palmerdabbelt@xxxxxxxxxx>

.