[PATCH 0/2] disable all pending tasklets during device removal

[PATCH 0/2] disable all pending tasklets during device removal

[Lorenzo Bianconi]

This patchset is based on 'add set_domain routine to reconfigure
mt76x2 dfs pd' patchset

Lorenzo Bianconi (2):
  mt76x2: init: disable tbtt tasklet during device cleanup
  mt76x2: dfs: disable dfs tasklet during device cleanup

 drivers/net/wireless/mediatek/mt76/mt76x2_dfs.c  | 6 ++++++
 drivers/net/wireless/mediatek/mt76/mt76x2_dfs.h  | 1 +
 drivers/net/wireless/mediatek/mt76/mt76x2_init.c | 2 ++
 3 files changed, 9 insertions(+)

-- 
2.13.6

[PATCH 1/2] mt76x2: init: disable tbtt tasklet during device cleanup

[Lorenzo Bianconi]

Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
---
 drivers/net/wireless/mediatek/mt76/mt76x2_init.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/net/wireless/mediatek/mt76/mt76x2_init.c b/drivers/net/wireless/mediatek/mt76/mt76x2_init.c
index 9f20a308ff76..66d4bfc70977 100644
--- a/drivers/net/wireless/mediatek/mt76/mt76x2_init.c
+++ b/drivers/net/wireless/mediatek/mt76/mt76x2_init.c
@@ -614,6 +614,7 @@ void mt76x2_stop_hardware(struct mt76x2_dev *dev)
 
 void mt76x2_cleanup(struct mt76x2_dev *dev)
 {
+	tasklet_disable(&dev->pre_tbtt_tasklet);
 	mt76x2_stop_hardware(dev);
 	mt76x2_dma_cleanup(dev);
 	mt76x2_mcu_cleanup(dev);
-- 
2.13.6

[PATCH 2/2] mt76x2: dfs: disable dfs tasklet during device cleanup

[Lorenzo Bianconi]

Define mt76x2_dfs_cleanup() routine since it will be helpful to add
the support for dfs sw detector

Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
---
 drivers/net/wireless/mediatek/mt76/mt76x2_dfs.c  | 6 ++++++
 drivers/net/wireless/mediatek/mt76/mt76x2_dfs.h  | 1 +
 drivers/net/wireless/mediatek/mt76/mt76x2_init.c | 1 +
 3 files changed, 8 insertions(+)

diff --git a/drivers/net/wireless/mediatek/mt76/mt76x2_dfs.c b/drivers/net/wireless/mediatek/mt76/mt76x2_dfs.c
index f936dc9a5476..19ac6edd672a 100644
--- a/drivers/net/wireless/mediatek/mt76/mt76x2_dfs.c
+++ b/drivers/net/wireless/mediatek/mt76/mt76x2_dfs.c
@@ -504,3 +504,9 @@ void mt76x2_dfs_set_domain(struct mt76x2_dev *dev,
 	}
 }
 
+void mt76x2_dfs_cleanup(struct mt76x2_dev *dev)
+{
+	struct mt76x2_dfs_pattern_detector *dfs_pd = &dev->dfs_pd;
+
+	tasklet_disable(&dfs_pd->dfs_tasklet);
+}
diff --git a/drivers/net/wireless/mediatek/mt76/mt76x2_dfs.h b/drivers/net/wireless/mediatek/mt76/mt76x2_dfs.h
index 8dbc783cc6bc..f70220615bbc 100644
--- a/drivers/net/wireless/mediatek/mt76/mt76x2_dfs.h
+++ b/drivers/net/wireless/mediatek/mt76/mt76x2_dfs.h
@@ -74,6 +74,7 @@ struct mt76x2_dfs_pattern_detector {
 };
 
 void mt76x2_dfs_init_params(struct mt76x2_dev *dev);
+void mt76x2_dfs_cleanup(struct mt76x2_dev *dev);
 void mt76x2_dfs_init_detector(struct mt76x2_dev *dev);
 void mt76x2_dfs_adjust_agc(struct mt76x2_dev *dev);
 void mt76x2_dfs_set_domain(struct mt76x2_dev *dev,
diff --git a/drivers/net/wireless/mediatek/mt76/mt76x2_init.c b/drivers/net/wireless/mediatek/mt76/mt76x2_init.c
index 66d4bfc70977..2bf861058658 100644
--- a/drivers/net/wireless/mediatek/mt76/mt76x2_init.c
+++ b/drivers/net/wireless/mediatek/mt76/mt76x2_init.c
@@ -618,6 +618,7 @@ void mt76x2_cleanup(struct mt76x2_dev *dev)
 	mt76x2_stop_hardware(dev);
 	mt76x2_dma_cleanup(dev);
 	mt76x2_mcu_cleanup(dev);
+	mt76x2_dfs_cleanup(dev);
 }
 
 struct mt76x2_dev *mt76x2_alloc_device(struct device *pdev)
-- 
2.13.6

Re: [PATCH 1/2] mt76x2: init: disable tbtt tasklet during device cleanup

[Kalle Valo]

Lorenzo Bianconi <lorenzo.bianconi@redhat.com> writes:

> Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>

No empty commit logs, please. At least the commit log should answer the
question "Why?". For example, are you fixing a reported bug or are you
fixing is a theoretical issue you found during code review? What kind of
symtomps does this bug have?

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches#commit_log_does_not_answer_why

-- 
Kalle Valo

Re: [PATCH 1/2] mt76x2: init: disable tbtt tasklet during device cleanup

[Lorenzo Bianconi]

> Lorenzo Bianconi <lorenzo.bianconi@redhat.com> writes:
>
>> Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
>
> No empty commit logs, please. At least the commit log should answer the
> question "Why?". For example, are you fixing a reported bug or are you
> fixing is a theoretical issue you found during code review? What kind of
> symtomps does this bug have?
>
> https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches#commit_log_does_not_answer_why

Ack, I will wait for some feedbacks and then I will respin a v2 with a
proper commit log. Thanks.
Regards,

Lorenzo

>
> --
> Kalle Valo