From: Lorenzo Bianconi <lorenzo@kernel.org>
To: nbd@nbd.name
Cc: linux-wireless@vger.kernel.org, lorenzo.bianconi@redhat.com,
sean.wang@mediatek.com
Subject: [PATCH 01/19] mt76: mt7921: fix a race between mt7921_mcu_drv_pmctrl and mt7921_mcu_fw_pmctrl
Date: Sun, 18 Apr 2021 18:45:27 +0200 [thread overview]
Message-ID: <dee187bcde4b0410e09061e8ad786c3bd769497b.1618763001.git.lorenzo@kernel.org> (raw)
In-Reply-To: <cover.1618763000.git.lorenzo@kernel.org>
Introduce a mutex in order to avoid a race between mt7921_mcu_drv_pmctrl and
mt7921_mcu_fw_pmctrl routines since they are run two independent works
Fixes: 1d8efc741df8 ("mt76: mt7921: introduce Runtime PM support")
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
---
.../net/wireless/mediatek/mt76/mt76_connac.h | 1 +
.../net/wireless/mediatek/mt76/mt7921/init.c | 1 +
.../net/wireless/mediatek/mt76/mt7921/mcu.c | 34 +++++++++++++------
3 files changed, 26 insertions(+), 10 deletions(-)
diff --git a/drivers/net/wireless/mediatek/mt76/mt76_connac.h b/drivers/net/wireless/mediatek/mt76/mt76_connac.h
index b811f3c410a1..2b31c9794e92 100644
--- a/drivers/net/wireless/mediatek/mt76/mt76_connac.h
+++ b/drivers/net/wireless/mediatek/mt76/mt76_connac.h
@@ -54,6 +54,7 @@ struct mt76_connac_pm {
struct work_struct wake_work;
struct completion wake_cmpl;
+ struct mutex mutex;
struct delayed_work ps_work;
unsigned long last_activity;
diff --git a/drivers/net/wireless/mediatek/mt76/mt7921/init.c b/drivers/net/wireless/mediatek/mt76/mt7921/init.c
index eab6e2dcdb96..0b8a5a7f4362 100644
--- a/drivers/net/wireless/mediatek/mt76/mt7921/init.c
+++ b/drivers/net/wireless/mediatek/mt76/mt7921/init.c
@@ -222,6 +222,7 @@ int mt7921_register_device(struct mt7921_dev *dev)
INIT_DELAYED_WORK(&dev->pm.ps_work, mt7921_pm_power_save_work);
INIT_WORK(&dev->pm.wake_work, mt7921_pm_wake_work);
+ mutex_init(&dev->pm.mutex);
init_completion(&dev->pm.wake_cmpl);
spin_lock_init(&dev->pm.txq_lock);
set_bit(MT76_STATE_PM, &dev->mphy.state);
diff --git a/drivers/net/wireless/mediatek/mt76/mt7921/mcu.c b/drivers/net/wireless/mediatek/mt76/mt7921/mcu.c
index 14ba856de0b6..ea00f6b6af56 100644
--- a/drivers/net/wireless/mediatek/mt76/mt7921/mcu.c
+++ b/drivers/net/wireless/mediatek/mt76/mt7921/mcu.c
@@ -1267,9 +1267,11 @@ int mt7921_mcu_set_bss_pm(struct mt7921_dev *dev, struct ieee80211_vif *vif,
int mt7921_mcu_drv_pmctrl(struct mt7921_dev *dev)
{
struct mt76_phy *mphy = &dev->mt76.phy;
- int i;
+ int i, err = 0;
- if (!test_and_clear_bit(MT76_STATE_PM, &mphy->state))
+ mutex_lock(&dev->pm.mutex);
+
+ if (!test_bit(MT76_STATE_PM, &mphy->state))
goto out;
for (i = 0; i < MT7921_DRV_OWN_RETRY_COUNT; i++) {
@@ -1281,23 +1283,30 @@ int mt7921_mcu_drv_pmctrl(struct mt7921_dev *dev)
if (i == MT7921_DRV_OWN_RETRY_COUNT) {
dev_err(dev->mt76.dev, "driver own failed\n");
- mt7921_reset(&dev->mt76);
- return -EIO;
+ err = -EIO;
+ goto out;
}
+ clear_bit(MT76_STATE_PM, &mphy->state);
out:
dev->pm.last_activity = jiffies;
+ mutex_unlock(&dev->pm.mutex);
- return 0;
+ if (err)
+ mt7921_reset(&dev->mt76);
+
+ return err;
}
int mt7921_mcu_fw_pmctrl(struct mt7921_dev *dev)
{
struct mt76_phy *mphy = &dev->mt76.phy;
- int i;
+ int i, err = 0;
+
+ mutex_lock(&dev->pm.mutex);
if (test_and_set_bit(MT76_STATE_PM, &mphy->state))
- return 0;
+ goto out;
for (i = 0; i < MT7921_DRV_OWN_RETRY_COUNT; i++) {
mt76_wr(dev, MT_CONN_ON_LPCTL, PCIE_LPCR_HOST_SET_OWN);
@@ -1308,11 +1317,16 @@ int mt7921_mcu_fw_pmctrl(struct mt7921_dev *dev)
if (i == MT7921_DRV_OWN_RETRY_COUNT) {
dev_err(dev->mt76.dev, "firmware own failed\n");
- mt7921_reset(&dev->mt76);
- return -EIO;
+ clear_bit(MT76_STATE_PM, &mphy->state);
+ err = -EIO;
}
+out:
+ mutex_unlock(&dev->pm.mutex);
- return 0;
+ if (err)
+ mt7921_reset(&dev->mt76);
+
+ return err;
}
void
--
2.30.2
next prev parent reply other threads:[~2021-04-18 16:45 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-04-18 16:45 [PATCH 00/19] mt76: improve runtime-pm support Lorenzo Bianconi
2021-04-18 16:45 ` Lorenzo Bianconi [this message]
2021-04-18 16:45 ` [PATCH 02/19] mt76: mt7663: fix a race between mt7615_mcu_drv_pmctrl and mt7615_mcu_fw_pmctrl Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 03/19] mt76: connac: introduce wake counter for fw_pmctrl synchronization Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 04/19] mt76: mt7921: rely on mt76_connac_pm_ref/mt76_connac_pm_unref in tx path Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 05/19] mt76: mt7663: " Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 06/19] mt76: dma: add the capability to define a custom rx napi poll routine Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 07/19] mt76: mt7921: rely on mt76_connac_pm_ref/mt76_connac_pm_unref in tx/rx napi Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 08/19] mt76: mt7663: " Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 09/19] mt76: connac: unschedule ps_work in mt76_connac_pm_wake Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 10/19] mt76: connac: check wake refcount in mcu_fw_pmctrl Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 11/19] mt76: connac: remove MT76_STATE_PM in mac_tx_free Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 12/19] mt76: mt7921: get rid of useless MT76_STATE_PM in mt7921_mac_work Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 13/19] mt76: connac: alaways wake the device before scanning Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 14/19] mt76: mt7615: rely on pm refcounting in mt7615_led_set_config Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 15/19] mt76: connac: do not run mt76_txq_schedule_all directly Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 16/19] mt76: connac: use waitqueue for runtime-pm Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 17/19] mt76: remove MT76_STATE_PM in tx path Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 18/19] mt76: mt7921: add awake and doze time accounting Lorenzo Bianconi
2021-04-18 16:45 ` [PATCH 19/19] mt76: mt7921: enable sw interrupts Lorenzo Bianconi
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=dee187bcde4b0410e09061e8ad786c3bd769497b.1618763001.git.lorenzo@kernel.org \
--to=lorenzo@kernel.org \
--cc=linux-wireless@vger.kernel.org \
--cc=lorenzo.bianconi@redhat.com \
--cc=nbd@nbd.name \
--cc=sean.wang@mediatek.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).