[PATCH net-next 0/5] net: ipa: kill off ipa_clock_get()

[PATCH net-next 0/5] net: ipa: kill off ipa_clock_get()

[Alex Elder]

This series replaces the remaining uses of ipa_clock_get() with
calls to pm_runtime_get_sync() instead.  It replaces all calls to
ipa_clock_put() with calls to pm_runtime_put().

This completes the preparation for enabling automated suspend under
the control of the power management core code.  The next patch (in
an upcoming series) enables that.  Then the "ipa_clock" files and
symbols will switch to using an "ipa_power" naming convention instead.

Additional info

It is possible for pm_runtime_get_sync() to return an error.  There
are really three cases, identified by return value:
  - 1, meaning power was already active
  - 0, meaning power was not previously active, but is now
  - EACCES, meaning runtime PM is disabled
One additional case is EINVAL, meaning a previous suspend or resume
(or idle) call returned an error.  But we have always assumed this
won't happen (we previously didn't even check for an error).

But because we use pm_runtime_force_suspend() to implement system
suspend, there's a chance we'd get an EACCES error (the first thing
that function does is disable runtime suspend).  Individual patches
explain what happens in that case, but generally we just accept that
it could be an unlikely problem (occurring only at startup time).

Similarly, pm_runtime_put() could return an error.  There too, we
ignore EINVAL, assuming the IPA suspend and resume operations won't
produce an error.  EBUSY and EPERM are not applicable, EAGAIN is not
expected (and harmless).  We should never get EACCES (runtime
suspend disabled), because pm_runtime_put() calls match prior
pm_runtime_get_sync() calls, and a system suspend will not be
started while a runtime suspend or resume is underway.  In summary,
the value returned from pm_runtime_put() is not meaningful, so we
explicitly ignore it.

					-Alex

Alex Elder (5):
  net: ipa: don't use ipa_clock_get() in "ipa_main.c"
  net: ipa: don't use ipa_clock_get() in "ipa_smp2p.c"
  net: ipa: don't use ipa_clock_get() in "ipa_uc.c"
  net: ipa: don't use ipa_clock_get() in "ipa_modem.c"
  net: ipa: kill ipa_clock_get()

 drivers/net/ipa/ipa_clock.c     | 17 --------------
 drivers/net/ipa/ipa_clock.h     | 24 --------------------
 drivers/net/ipa/ipa_interrupt.c | 14 ++++++------
 drivers/net/ipa/ipa_main.c      | 21 ++++++++---------
 drivers/net/ipa/ipa_modem.c     | 40 +++++++++++++++++++--------------
 drivers/net/ipa/ipa_smp2p.c     | 19 +++++++++-------
 drivers/net/ipa/ipa_uc.c        | 22 ++++++++++--------
 7 files changed, 65 insertions(+), 92 deletions(-)

-- 
2.27.0

[PATCH net-next 1/5] net: ipa: don't use ipa_clock_get() in "ipa_main.c"

[Alex Elder]

We need the hardware to be powered starting at the config stage of
initialization when the IPA driver probes.  And we need it powered
when the driver is removed, at least until the deconfig stage has
completed.

Replace callers of ipa_clock_get() in ipa_probe() and ipa_exit(),
calling pm_runtime_get_sync() instead.  Replace the corresponding
callers of ipa_clock_put(), calling pm_runtime_put() instead.

The only error we expect when getting power would occur when the
system is suspended.  The ->probe and ->remove driver callbacks
won't be called when suspended, so issue a WARN() call if an error
is seen getting power.

Signed-off-by: Alex Elder <elder@linaro.org>
---
 drivers/net/ipa/ipa_main.c | 21 +++++++++++----------
 1 file changed, 11 insertions(+), 10 deletions(-)

diff --git a/drivers/net/ipa/ipa_main.c b/drivers/net/ipa/ipa_main.c
index 69fa4b3120fd3..3969aef6c4370 100644
--- a/drivers/net/ipa/ipa_main.c
+++ b/drivers/net/ipa/ipa_main.c
@@ -15,6 +15,7 @@
 #include <linux/of.h>
 #include <linux/of_device.h>
 #include <linux/of_address.h>
+#include <linux/pm_runtime.h>
 #include <linux/qcom_scm.h>
 #include <linux/soc/qcom/mdt_loader.h>
 
@@ -737,13 +738,13 @@ static int ipa_probe(struct platform_device *pdev)
 		goto err_table_exit;
 
 	/* The clock needs to be active for config and setup */
-	ret = ipa_clock_get(ipa);
+	ret = pm_runtime_get_sync(dev);
 	if (WARN_ON(ret < 0))
-		goto err_clock_put;
+		goto err_power_put;
 
 	ret = ipa_config(ipa, data);
 	if (ret)
-		goto err_clock_put;
+		goto err_power_put;
 
 	dev_info(dev, "IPA driver initialized");
 
@@ -765,14 +766,14 @@ static int ipa_probe(struct platform_device *pdev)
 	if (ret)
 		goto err_deconfig;
 done:
-	(void)ipa_clock_put(ipa);
+	(void)pm_runtime_put(dev);
 
 	return 0;
 
 err_deconfig:
 	ipa_deconfig(ipa);
-err_clock_put:
-	(void)ipa_clock_put(ipa);
+err_power_put:
+	(void)pm_runtime_put(dev);
 	ipa_modem_exit(ipa);
 err_table_exit:
 	ipa_table_exit(ipa);
@@ -798,9 +799,9 @@ static int ipa_remove(struct platform_device *pdev)
 	struct ipa_clock *clock = ipa->clock;
 	int ret;
 
-	ret = ipa_clock_get(ipa);
+	ret = pm_runtime_get_sync(&pdev->dev);
 	if (WARN_ON(ret < 0))
-		goto out_clock_put;
+		goto out_power_put;
 
 	if (ipa->setup_complete) {
 		ret = ipa_modem_stop(ipa);
@@ -816,8 +817,8 @@ static int ipa_remove(struct platform_device *pdev)
 	}
 
 	ipa_deconfig(ipa);
-out_clock_put:
-	(void)ipa_clock_put(ipa);
+out_power_put:
+	(void)pm_runtime_put(&pdev->dev);
 
 	ipa_modem_exit(ipa);
 	ipa_table_exit(ipa);
-- 
2.27.0

[PATCH net-next 2/5] net: ipa: don't use ipa_clock_get() in "ipa_smp2p.c"

[Alex Elder]

If the "modem-init" Device Tree property is present for a platform,
the modem performs early IPA hardware initialization, and signals
this is complete with an "ipa-setup-ready" SMP2P interrupt.  This
triggers a call to ipa_setup(), which requires the hardware to be
powered.

Replace the call to ipa_clock_get() in this case with a call to
pm_runtime_get_sync().  And replace the corresponding calls to
ipa_clock_put() with calls to pm_runtime_put() instead.

There is a chance we get an error when taking this power reference.
This is an unlikely scenario, where system suspend is initiated just
before the modem signals it has finished initializing the IPA
hardware.  For now we'll just accept that this could occur, and
report it if it does.

Signed-off-by: Alex Elder <elder@linaro.org>
---
 drivers/net/ipa/ipa_smp2p.c | 19 +++++++++++--------
 1 file changed, 11 insertions(+), 8 deletions(-)

diff --git a/drivers/net/ipa/ipa_smp2p.c b/drivers/net/ipa/ipa_smp2p.c
index 04b977cf91593..f6e2061cd391c 100644
--- a/drivers/net/ipa/ipa_smp2p.c
+++ b/drivers/net/ipa/ipa_smp2p.c
@@ -16,7 +16,6 @@
 #include "ipa_smp2p.h"
 #include "ipa.h"
 #include "ipa_uc.h"
-#include "ipa_clock.h"
 
 /**
  * DOC: IPA SMP2P communication with the modem
@@ -153,6 +152,7 @@ static void ipa_smp2p_panic_notifier_unregister(struct ipa_smp2p *smp2p)
 static irqreturn_t ipa_smp2p_modem_setup_ready_isr(int irq, void *dev_id)
 {
 	struct ipa_smp2p *smp2p = dev_id;
+	struct device *dev;
 	int ret;
 
 	mutex_lock(&smp2p->mutex);
@@ -161,17 +161,20 @@ static irqreturn_t ipa_smp2p_modem_setup_ready_isr(int irq, void *dev_id)
 		goto out_mutex_unlock;
 	smp2p->disabled = true;		/* If any others arrive, ignore them */
 
-	/* The clock needs to be active for setup */
-	ret = ipa_clock_get(smp2p->ipa);
-	if (WARN_ON(ret < 0))
-		goto out_clock_put;
+	/* Power needs to be active for setup */
+	dev = &smp2p->ipa->pdev->dev;
+	ret = pm_runtime_get_sync(dev);
+	if (ret < 0) {
+		dev_err(dev, "error %d getting power for setup\n", ret);
+		goto out_power_put;
+	}
 
 	/* An error here won't cause driver shutdown, so warn if one occurs */
 	ret = ipa_setup(smp2p->ipa);
 	WARN(ret != 0, "error %d from ipa_setup()\n", ret);
 
-out_clock_put:
-	(void)ipa_clock_put(smp2p->ipa);
+out_power_put:
+	(void)pm_runtime_put(dev);
 out_mutex_unlock:
 	mutex_unlock(&smp2p->mutex);
 
@@ -211,7 +214,7 @@ static void ipa_smp2p_clock_release(struct ipa *ipa)
 	if (!ipa->smp2p->clock_on)
 		return;
 
-	(void)ipa_clock_put(ipa);
+	(void)pm_runtime_put(&ipa->pdev->dev);
 	ipa->smp2p->clock_on = false;
 }
 
-- 
2.27.0

[PATCH net-next 3/5] net: ipa: don't use ipa_clock_get() in "ipa_uc.c"

[Alex Elder]

Replace the ipa_clock_get() call in ipa_uc_clock() when taking the
"proxy" clock reference for the microcontroller with a call to
pm_runtime_get_sync().  Replace calls of ipa_clock_put() for the
microcontroller with pm_runtime_put() calls instead.

There is a chance we get an error when taking the microcontroller
power reference.  This is an unlikely scenario, where system suspend
is initiated just before we learn the modem is booting.  For now
we'll just accept that this could occur, and report it if it does.

Signed-off-by: Alex Elder <elder@linaro.org>
---
 drivers/net/ipa/ipa_uc.c | 22 +++++++++++++---------
 1 file changed, 13 insertions(+), 9 deletions(-)

diff --git a/drivers/net/ipa/ipa_uc.c b/drivers/net/ipa/ipa_uc.c
index 9c8818c390731..a0bdd25b65b4f 100644
--- a/drivers/net/ipa/ipa_uc.c
+++ b/drivers/net/ipa/ipa_uc.c
@@ -7,9 +7,9 @@
 #include <linux/types.h>
 #include <linux/io.h>
 #include <linux/delay.h>
+#include <linux/pm_runtime.h>
 
 #include "ipa.h"
-#include "ipa_clock.h"
 #include "ipa_uc.h"
 
 /**
@@ -154,7 +154,7 @@ static void ipa_uc_response_hdlr(struct ipa *ipa, enum ipa_irq_id irq_id)
 	case IPA_UC_RESPONSE_INIT_COMPLETED:
 		if (ipa->uc_clocked) {
 			ipa->uc_loaded = true;
-			(void)ipa_clock_put(ipa);
+			(void)pm_runtime_put(dev);
 			ipa->uc_clocked = false;
 		} else {
 			dev_warn(dev, "unexpected init_completed response\n");
@@ -182,25 +182,29 @@ void ipa_uc_deconfig(struct ipa *ipa)
 	ipa_interrupt_remove(ipa->interrupt, IPA_IRQ_UC_1);
 	ipa_interrupt_remove(ipa->interrupt, IPA_IRQ_UC_0);
 	if (ipa->uc_clocked)
-		(void)ipa_clock_put(ipa);
+		(void)pm_runtime_put(&ipa->pdev->dev);
 }
 
 /* Take a proxy clock reference for the microcontroller */
 void ipa_uc_clock(struct ipa *ipa)
 {
 	static bool already;
+	struct device *dev;
 	int ret;
 
 	if (already)
 		return;
 	already = true;		/* Only do this on first boot */
 
-	/* This clock reference dropped in ipa_uc_response_hdlr() above */
-	ret = ipa_clock_get(ipa);
-	if (WARN(ret < 0, "error %d getting proxy clock\n", ret))
-		(void)ipa_clock_put(ipa);
-
-	ipa->uc_clocked = ret >= 0;
+	/* This power reference dropped in ipa_uc_response_hdlr() above */
+	dev = &ipa->pdev->dev;
+	ret = pm_runtime_get_sync(dev);
+	if (ret < 0) {
+		pm_runtime_put_noidle(dev);
+		dev_err(dev, "error %d getting proxy power\n", ret);
+	} else {
+		ipa->uc_clocked = true;
+	}
 }
 
 /* Send a command to the microcontroller */
-- 
2.27.0

[PATCH net-next 4/5] net: ipa: don't use ipa_clock_get() in "ipa_modem.c"

[Alex Elder]

When we open or close the modem network device we need to ensure the
hardware is powered.  Replace the callers of ipa_clock_get() found
in ipa_open() and ipa_stop() with calls to pm_runtime_get_sync().
If an error is returned, simply return that error to the caller
(without any error or warning message).  This could conceivably
occur if the function was called while the system was suspended,
but that really shouldn't happen.  Replace corresponding calls to
ipa_clock_put() with pm_runtime_put() also.

If the modem crashes we also need to ensure the hardware is powered
to recover.  If getting power returns an error there's not much we
can do, but at least report the error.  (Ideally the remoteproc SSR
code would ensure the AP was not suspended when it sends the
notification, but that is not (yet) the case.)

Signed-off-by: Alex Elder <elder@linaro.org>
---
 drivers/net/ipa/ipa_modem.c | 40 +++++++++++++++++++++----------------
 1 file changed, 23 insertions(+), 17 deletions(-)

diff --git a/drivers/net/ipa/ipa_modem.c b/drivers/net/ipa/ipa_modem.c
index c8724af935b85..7e05b22d6e18b 100644
--- a/drivers/net/ipa/ipa_modem.c
+++ b/drivers/net/ipa/ipa_modem.c
@@ -49,15 +49,17 @@ static int ipa_open(struct net_device *netdev)
 {
 	struct ipa_priv *priv = netdev_priv(netdev);
 	struct ipa *ipa = priv->ipa;
+	struct device *dev;
 	int ret;
 
-	ret = ipa_clock_get(ipa);
-	if (WARN_ON(ret < 0))
-		goto err_clock_put;
+	dev = &ipa->pdev->dev;
+	ret = pm_runtime_get_sync(dev);
+	if (ret < 0)
+		goto err_power_put;
 
 	ret = ipa_endpoint_enable_one(ipa->name_map[IPA_ENDPOINT_AP_MODEM_TX]);
 	if (ret)
-		goto err_clock_put;
+		goto err_power_put;
 
 	ret = ipa_endpoint_enable_one(ipa->name_map[IPA_ENDPOINT_AP_MODEM_RX]);
 	if (ret)
@@ -65,14 +67,14 @@ static int ipa_open(struct net_device *netdev)
 
 	netif_start_queue(netdev);
 
-	(void)ipa_clock_put(ipa);
+	(void)pm_runtime_put(dev);
 
 	return 0;
 
 err_disable_tx:
 	ipa_endpoint_disable_one(ipa->name_map[IPA_ENDPOINT_AP_MODEM_TX]);
-err_clock_put:
-	(void)ipa_clock_put(ipa);
+err_power_put:
+	(void)pm_runtime_put(dev);
 
 	return ret;
 }
@@ -82,18 +84,20 @@ static int ipa_stop(struct net_device *netdev)
 {
 	struct ipa_priv *priv = netdev_priv(netdev);
 	struct ipa *ipa = priv->ipa;
+	struct device *dev;
 	int ret;
 
-	ret = ipa_clock_get(ipa);
-	if (WARN_ON(ret < 0))
-		goto out_clock_put;
+	dev = &ipa->pdev->dev;
+	ret = pm_runtime_get_sync(dev);
+	if (ret < 0)
+		goto out_power_put;
 
 	netif_stop_queue(netdev);
 
 	ipa_endpoint_disable_one(ipa->name_map[IPA_ENDPOINT_AP_MODEM_RX]);
 	ipa_endpoint_disable_one(ipa->name_map[IPA_ENDPOINT_AP_MODEM_TX]);
-out_clock_put:
-	(void)ipa_clock_put(ipa);
+out_power_put:
+	(void)pm_runtime_put(dev);
 
 	return 0;
 }
@@ -359,9 +363,11 @@ static void ipa_modem_crashed(struct ipa *ipa)
 	struct device *dev = &ipa->pdev->dev;
 	int ret;
 
-	ret = ipa_clock_get(ipa);
-	if (WARN_ON(ret < 0))
-		goto out_clock_put;
+	ret = pm_runtime_get_sync(dev);
+	if (ret < 0) {
+		dev_err(dev, "error %d getting power to handle crash\n", ret);
+		goto out_power_put;
+	}
 
 	ipa_endpoint_modem_pause_all(ipa, true);
 
@@ -388,8 +394,8 @@ static void ipa_modem_crashed(struct ipa *ipa)
 	if (ret)
 		dev_err(dev, "error %d zeroing modem memory regions\n", ret);
 
-out_clock_put:
-	(void)ipa_clock_put(ipa);
+out_power_put:
+	(void)pm_runtime_put(dev);
 }
 
 static int ipa_modem_notify(struct notifier_block *nb, unsigned long action,
-- 
2.27.0

[PATCH net-next 5/5] net: ipa: kill ipa_clock_get()

[Alex Elder]

The only remaining user of the ipa_clock_{get,put}() interface is
ipa_isr_thread().  Replace calls to ipa_clock_get() there calling
pm_runtime_get_sync() instead.  And call pm_runtime_put() there
rather than ipa_clock_put().  Warn if we ever get an error.

With that, we can get rid of ipa_clock_get() and ipa_clock_put().

Signed-off-by: Alex Elder <elder@linaro.org>
---
 drivers/net/ipa/ipa_clock.c     | 17 -----------------
 drivers/net/ipa/ipa_clock.h     | 24 ------------------------
 drivers/net/ipa/ipa_interrupt.c | 14 +++++++-------
 3 files changed, 7 insertions(+), 48 deletions(-)

diff --git a/drivers/net/ipa/ipa_clock.c b/drivers/net/ipa/ipa_clock.c
index 8f25107c1f1e7..357be73a45834 100644
--- a/drivers/net/ipa/ipa_clock.c
+++ b/drivers/net/ipa/ipa_clock.c
@@ -266,23 +266,6 @@ static int ipa_runtime_idle(struct device *dev)
 	return -EAGAIN;
 }
 
-/* Get an IPA clock reference.  If the reference count is non-zero, it is
- * incremented and return is immediate.  Otherwise the IPA clock is
- * enabled.
- */
-int ipa_clock_get(struct ipa *ipa)
-{
-	return pm_runtime_get_sync(&ipa->pdev->dev);
-}
-
-/* Attempt to remove an IPA clock reference.  If this represents the
- * last reference, disable the IPA clock.
- */
-int ipa_clock_put(struct ipa *ipa)
-{
-	return pm_runtime_put(&ipa->pdev->dev);
-}
-
 static int ipa_suspend(struct device *dev)
 {
 	struct ipa *ipa = dev_get_drvdata(dev);
diff --git a/drivers/net/ipa/ipa_clock.h b/drivers/net/ipa/ipa_clock.h
index 5c53241336a1a..8c21a007c4375 100644
--- a/drivers/net/ipa/ipa_clock.h
+++ b/drivers/net/ipa/ipa_clock.h
@@ -52,28 +52,4 @@ struct ipa_clock *ipa_clock_init(struct device *dev,
  */
 void ipa_clock_exit(struct ipa_clock *clock);
 
-/**
- * ipa_clock_get() - Get an IPA clock reference
- * @ipa:	IPA pointer
- *
- * Return:	0 if clock started, 1 if clock already running, or a negative
- *		error code
- *
- * This call blocks if this is the first reference.  A reference is
- * taken even if an error occurs starting the IPA clock.
- */
-int ipa_clock_get(struct ipa *ipa);
-
-/**
- * ipa_clock_put() - Drop an IPA clock reference
- * @ipa:	IPA pointer
- *
- * Return:	0 if successful, or a negative error code
- *
- * This drops a clock reference.  If the last reference is being dropped,
- * the clock is stopped and RX endpoints are suspended.  This call will
- * not block unless the last reference is dropped.
- */
-int ipa_clock_put(struct ipa *ipa);
-
 #endif /* _IPA_CLOCK_H_ */
diff --git a/drivers/net/ipa/ipa_interrupt.c b/drivers/net/ipa/ipa_interrupt.c
index 934c14e066a0a..3fecaadb4a37e 100644
--- a/drivers/net/ipa/ipa_interrupt.c
+++ b/drivers/net/ipa/ipa_interrupt.c
@@ -21,9 +21,9 @@
 
 #include <linux/types.h>
 #include <linux/interrupt.h>
+#include <linux/pm_runtime.h>
 
 #include "ipa.h"
-#include "ipa_clock.h"
 #include "ipa_reg.h"
 #include "ipa_endpoint.h"
 #include "ipa_interrupt.h"
@@ -80,14 +80,16 @@ static irqreturn_t ipa_isr_thread(int irq, void *dev_id)
 	struct ipa_interrupt *interrupt = dev_id;
 	struct ipa *ipa = interrupt->ipa;
 	u32 enabled = interrupt->enabled;
+	struct device *dev;
 	u32 pending;
 	u32 offset;
 	u32 mask;
 	int ret;
 
-	ret = ipa_clock_get(ipa);
+	dev = &ipa->pdev->dev;
+	ret = pm_runtime_get_sync(dev);
 	if (WARN_ON(ret < 0))
-		goto out_clock_put;
+		goto out_power_put;
 
 	/* The status register indicates which conditions are present,
 	 * including conditions whose interrupt is not enabled.  Handle
@@ -108,15 +110,13 @@ static irqreturn_t ipa_isr_thread(int irq, void *dev_id)
 
 	/* If any disabled interrupts are pending, clear them */
 	if (pending) {
-		struct device *dev = &ipa->pdev->dev;
-
 		dev_dbg(dev, "clearing disabled IPA interrupts 0x%08x\n",
 			pending);
 		offset = ipa_reg_irq_clr_offset(ipa->version);
 		iowrite32(pending, ipa->reg_virt + offset);
 	}
-out_clock_put:
-	(void)ipa_clock_put(ipa);
+out_power_put:
+	(void)pm_runtime_put(dev);
 
 	return IRQ_HANDLED;
 }
-- 
2.27.0

Re: [PATCH net-next 0/5] net: ipa: kill off ipa_clock_get()

[patchwork-bot+netdevbpf]

Hello:

This series was applied to netdev/net-next.git (refs/heads/master):

On Thu, 19 Aug 2021 17:19:22 -0500 you wrote:
> This series replaces the remaining uses of ipa_clock_get() with
> calls to pm_runtime_get_sync() instead.  It replaces all calls to
> ipa_clock_put() with calls to pm_runtime_put().
> 
> This completes the preparation for enabling automated suspend under
> the control of the power management core code.  The next patch (in
> an upcoming series) enables that.  Then the "ipa_clock" files and
> symbols will switch to using an "ipa_power" naming convention instead.
> 
> [...]

Here is the summary with links:
  - [net-next,1/5] net: ipa: don't use ipa_clock_get() in "ipa_main.c"
    https://git.kernel.org/netdev/net-next/c/4c6a4da84431
  - [net-next,2/5] net: ipa: don't use ipa_clock_get() in "ipa_smp2p.c"
    https://git.kernel.org/netdev/net-next/c/c43adc75dc2d
  - [net-next,3/5] net: ipa: don't use ipa_clock_get() in "ipa_uc.c"
    https://git.kernel.org/netdev/net-next/c/799c5c24b7ac
  - [net-next,4/5] net: ipa: don't use ipa_clock_get() in "ipa_modem.c"
    https://git.kernel.org/netdev/net-next/c/724c2d743688
  - [net-next,5/5] net: ipa: kill ipa_clock_get()
    https://git.kernel.org/netdev/net-next/c/c3f115aa5e1b

You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html