Re: [PATCH V5 1/2] watchdog: imx_sc: Add i.MX system controller watchdog support
From: Guenter Roeck
Date: Mon Mar 04 2019 - 08:53:12 EST
On 3/3/19 7:22 PM, Anson Huang wrote:
Hi, Guenter
Best Regards!
Anson Huang
-----Original Message-----
From: Guenter Roeck [mailto:groeck7@xxxxxxxxx] On Behalf Of Guenter
Roeck
Sent: 2019年3月4日 10:46
To: Anson Huang <anson.huang@xxxxxxx>
Cc: catalin.marinas@xxxxxxx; will.deacon@xxxxxxx; wim@linux-
watchdog.org; shawnguo@xxxxxxxxxx; s.hauer@xxxxxxxxxxxxxx;
kernel@xxxxxxxxxxxxxx; festevam@xxxxxxxxx; Andy Gross
<andy.gross@xxxxxxxxxx>; heiko@xxxxxxxxx; horms+renesas@xxxxxxxxxxxx;
arnd@xxxxxxxx; olof@xxxxxxxxx; bjorn.andersson@xxxxxxxxxx;
jagan@xxxxxxxxxxxxxxxxxxxx; enric.balletbo@xxxxxxxxxxxxx;
marc.w.gonzalez@xxxxxxx; linux-arm-kernel@xxxxxxxxxxxxxxxxxxx; linux-
kernel@xxxxxxxxxxxxxxx; linux-watchdog@xxxxxxxxxxxxxxx; dl-linux-imx
<linux-imx@xxxxxxx>
Subject: Re: [PATCH V5 1/2] watchdog: imx_sc: Add i.MX system controller
watchdog support
On 3/3/19 5:32 PM, Anson Huang wrote:
Hi, Guenter
Best Regards!
Anson Huang
-----Original Message-----
From: Guenter Roeck [mailto:groeck7@xxxxxxxxx] On Behalf Of Guenter
Roeck
Sent: 2019年3月2日 2:32
To: Anson Huang <anson.huang@xxxxxxx>
Cc: catalin.marinas@xxxxxxx; will.deacon@xxxxxxx; wim@linux-
watchdog.org; shawnguo@xxxxxxxxxx; s.hauer@xxxxxxxxxxxxxx;
kernel@xxxxxxxxxxxxxx; festevam@xxxxxxxxx; Andy Gross
<andy.gross@xxxxxxxxxx>; heiko@xxxxxxxxx;
horms+renesas@xxxxxxxxxxxx;
arnd@xxxxxxxx; olof@xxxxxxxxx; bjorn.andersson@xxxxxxxxxx;
jagan@xxxxxxxxxxxxxxxxxxxx; enric.balletbo@xxxxxxxxxxxxx;
marc.w.gonzalez@xxxxxxx; linux-arm-kernel@xxxxxxxxxxxxxxxxxxx; linux-
kernel@xxxxxxxxxxxxxxx; linux-watchdog@xxxxxxxxxxxxxxx; dl-linux-imx
<linux-imx@xxxxxxx>
Subject: Re: [PATCH V5 1/2] watchdog: imx_sc: Add i.MX system
controller watchdog support
On Fri, Mar 01, 2019 at 06:35:31AM +0000, Anson Huang wrote:
i.MX8QXP is an ARMv8 SoC which has a Cortex-M4 system controller
inside, the system controller is in charge of controlling power,
clock and watchdog etc..
This patch adds i.MX system controller watchdog driver support,
watchdog operation needs to be done in secure EL3 mode via
ARM-Trusted-Firmware, using SMC call, CPU will trap into
ARM-Trusted-Firmware and then it will request system controller to
do watchdog operation via IPC.
Signed-off-by: Anson Huang <Anson.Huang@xxxxxxx>
---
Changes since V4:
- change the module build dependency as depends on IMX_SCU and
HAVE_ARM_SMCCC, as this
driver is ONLY for i.MX SoC with SCU inside and it uses ARM SMC call.
---
drivers/watchdog/Kconfig | 14 +++
drivers/watchdog/Makefile | 1 +
drivers/watchdog/imx_sc_wdt.c | 201
++++++++++++++++++++++++++++++++++++++++++
3 files changed, 216 insertions(+)
create mode 100644 drivers/watchdog/imx_sc_wdt.c
diff --git a/drivers/watchdog/Kconfig b/drivers/watchdog/Kconfig
index
65c3c42..a6bfa54 100644
--- a/drivers/watchdog/Kconfig
+++ b/drivers/watchdog/Kconfig
@@ -625,6 +625,20 @@ config IMX2_WDT
To compile this driver as a module, choose M here: the
module will be called imx2_wdt.
+config IMX_SC_WDT
+ tristate "IMX SC Watchdog"
+ depends on IMX_SCU
+ depends on HAVE_ARM_SMCCC
+ select WATCHDOG_CORE
+ help
+ This is the driver for the system controller watchdog
+ on the NXP i.MX SoCs with system controller inside.
+ If you have one of these processors and wish to have
+ watchdog support enabled, say Y, otherwise say N.
+
+ To compile this driver as a module, choose M here: the
+ module will be called imx_sc_wdt.
+
config UX500_WATCHDOG
tristate "ST-Ericsson Ux500 watchdog"
depends on MFD_DB8500_PRCMU
diff --git a/drivers/watchdog/Makefile b/drivers/watchdog/Makefile
index 4e78a8c..0c9da63 100644
--- a/drivers/watchdog/Makefile
+++ b/drivers/watchdog/Makefile
@@ -68,6 +68,7 @@ obj-$(CONFIG_NUC900_WATCHDOG) +=
nuc900_wdt.o
obj-$(CONFIG_TS4800_WATCHDOG) += ts4800_wdt.o
obj-$(CONFIG_TS72XX_WATCHDOG) += ts72xx_wdt.o
obj-$(CONFIG_IMX2_WDT) += imx2_wdt.o
+obj-$(CONFIG_IMX_SC_WDT) += imx_sc_wdt.o
obj-$(CONFIG_UX500_WATCHDOG) += ux500_wdt.o
obj-$(CONFIG_RETU_WATCHDOG) += retu_wdt.o
obj-$(CONFIG_BCM2835_WDT) += bcm2835_wdt.o diff --git
a/drivers/watchdog/imx_sc_wdt.c b/drivers/watchdog/imx_sc_wdt.c
new
file mode 100644 index 0000000..50b49b2
--- /dev/null
+++ b/drivers/watchdog/imx_sc_wdt.c
@@ -0,0 +1,201 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright 2018-2019 NXP.
+ */
+
+#include <linux/arm-smccc.h>
+#include <linux/io.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/of.h>
Should no longer be needed.
Correct, I will remove it.
+#include <linux/platform_device.h>
+#include <linux/reboot.h>
+#include <linux/watchdog.h>
+
+#define DEFAULT_TIMEOUT 60
+/*
+ * Software timer tick implemented in scfw side, support 10ms to
+0xffffffff ms
+ * in theory, but for normal case, 1s~128s is enough, you can
+change this max
+ * value in case it's not enough.
+ */
+#define MAX_TIMEOUT 128
+
+#define IMX_SIP_TIMER 0xC2000002
+#define IMX_SIP_TIMER_START_WDOG 0x01
+#define IMX_SIP_TIMER_STOP_WDOG 0x02
+#define IMX_SIP_TIMER_SET_WDOG_ACT 0x03
+#define IMX_SIP_TIMER_PING_WDOG 0x04
+#define IMX_SIP_TIMER_SET_TIMEOUT_WDOG 0x05
+#define IMX_SIP_TIMER_GET_WDOG_STAT 0x06
+#define IMX_SIP_TIMER_SET_PRETIME_WDOG 0x07
+
+#define SC_TIMER_WDOG_ACTION_PARTITION 0
+
+static bool nowayout = WATCHDOG_NOWAYOUT;
module_param(nowayout,
+bool, 0000); MODULE_PARM_DESC(nowayout, "Watchdog cannot be
stopped
+once started (default="
+ __MODULE_STRING(WATCHDOG_NOWAYOUT) ")");
+
+static unsigned int timeout = DEFAULT_TIMEOUT;
+module_param(timeout, uint, 0000); MODULE_PARM_DESC(timeout,
+"Watchdog timeout in
seconds
+(default="
+ __MODULE_STRING(DEFAULT_TIMEOUT) ")");
+
+static struct platform_device *imx_sc_wdt_pdev;
+
+static int imx_sc_wdt_ping(struct watchdog_device *wdog) {
+ struct arm_smccc_res res;
+
+ arm_smccc_smc(IMX_SIP_TIMER, IMX_SIP_TIMER_PING_WDOG,
+ 0, 0, 0, 0, 0, 0, &res);
+
+ return 0;
+}
+
+static int imx_sc_wdt_start(struct watchdog_device *wdog) {
+ struct arm_smccc_res res;
+
+ arm_smccc_smc(IMX_SIP_TIMER, IMX_SIP_TIMER_START_WDOG,
+ 0, 0, 0, 0, 0, 0, &res);
+ if (res.a0)
+ return -EACCES;
+
+ arm_smccc_smc(IMX_SIP_TIMER, IMX_SIP_TIMER_SET_WDOG_ACT,
+ SC_TIMER_WDOG_ACTION_PARTITION,
+ 0, 0, 0, 0, 0, &res);
+ return res.a0 ? -EACCES : 0;
+}
+
+static int imx_sc_wdt_stop(struct watchdog_device *wdog) {
+ struct arm_smccc_res res;
+
+ arm_smccc_smc(IMX_SIP_TIMER, IMX_SIP_TIMER_STOP_WDOG,
+ 0, 0, 0, 0, 0, 0, &res);
+
+ return res.a0 ? -EACCES : 0;
+}
+
+static int imx_sc_wdt_set_timeout(struct watchdog_device *wdog,
+ unsigned int timeout)
+{
+ struct arm_smccc_res res;
+
+ wdog->timeout = timeout;
+ arm_smccc_smc(IMX_SIP_TIMER,
IMX_SIP_TIMER_SET_TIMEOUT_WDOG,
+ timeout * 1000, 0, 0, 0, 0, 0, &res);
+
+ return res.a0 ? -EACCES : 0;
+}
+
+static const struct watchdog_ops imx_sc_wdt_ops = {
+ .owner = THIS_MODULE,
+ .start = imx_sc_wdt_start,
+ .stop = imx_sc_wdt_stop,
+ .ping = imx_sc_wdt_ping,
+ .set_timeout = imx_sc_wdt_set_timeout, };
+
+static const struct watchdog_info imx_sc_wdt_info = {
+ .identity = "i.MX SC watchdog timer",
+ .options = WDIOF_SETTIMEOUT | WDIOF_KEEPALIVEPING |
+ WDIOF_MAGICCLOSE | WDIOF_PRETIMEOUT, };
+
+static int imx_sc_wdt_probe(struct platform_device *pdev) {
+ struct watchdog_device *imx_sc_wdd;
+ int ret;
+
+ imx_sc_wdd = devm_kzalloc(&pdev->dev, sizeof(*imx_sc_wdd),
GFP_KERNEL);
+ if (!imx_sc_wdd)
+ return -ENOMEM;
+
+ platform_set_drvdata(pdev, imx_sc_wdd);
+
+ imx_sc_wdd->info = &imx_sc_wdt_info;
+ imx_sc_wdd->ops = &imx_sc_wdt_ops;
+ imx_sc_wdd->min_timeout = 1;
+ imx_sc_wdd->max_timeout = MAX_TIMEOUT;
+ imx_sc_wdd->parent = &pdev->dev;
+ imx_sc_wdd->timeout = DEFAULT_TIMEOUT;
+
+ ret = watchdog_init_timeout(imx_sc_wdd, timeout, &pdev->dev);
+ if (ret)
+ dev_warn(&pdev->dev, "Failed to set timeout value, using
+default\n");
+
+ watchdog_stop_on_reboot(imx_sc_wdd);
+ watchdog_stop_on_unregister(imx_sc_wdd);
+
+ ret = devm_watchdog_register_device(&pdev->dev, imx_sc_wdd);
+ if (ret) {
+ dev_err(&pdev->dev, "Failed to register watchdog device\n");
+ return ret;
+ }
+
+ return 0;
+}
+
+static int __maybe_unused imx_sc_wdt_suspend(struct device *dev) {
+ struct watchdog_device *imx_sc_wdd = dev_get_drvdata(dev);
+
+ if (watchdog_active(imx_sc_wdd))
+ imx_sc_wdt_stop(imx_sc_wdd);
+
+ return 0;
+}
+
+static int __maybe_unused imx_sc_wdt_resume(struct device *dev) {
+ struct watchdog_device *imx_sc_wdd = dev_get_drvdata(dev);
+
+ if (watchdog_active(imx_sc_wdd))
+ imx_sc_wdt_start(imx_sc_wdd);
+
+ return 0;
+}
+
+static SIMPLE_DEV_PM_OPS(imx_sc_wdt_pm_ops,
+ imx_sc_wdt_suspend, imx_sc_wdt_resume);
+
+static struct platform_driver imx_sc_wdt_driver = {
+ .probe = imx_sc_wdt_probe,
+ .driver = {
+ .name = "imx-sc-wdt",
+ .pm = &imx_sc_wdt_pm_ops,
+ },
+};
+
+static int __init imx_sc_wdt_init(void) {
+ int ret;
+
+ ret = platform_driver_register(&imx_sc_wdt_driver);
+ if (ret)
+ return ret;
+
+ imx_sc_wdt_pdev = platform_device_register_simple("imx-sc-wdt", -
1, NULL, 0);
+ if (IS_ERR(imx_sc_wdt_pdev)) {
+ platform_driver_unregister(&imx_sc_wdt_driver);
+ return PTR_ERR(imx_sc_wdt_pdev);
+ }
I just realized what you are doing here. So the watchdog will always
be instantiated if/when the module is loaded. I don't think that was
the idea, and it seems to be risky. What happens if someone loads the
module on a system where the watchdog is not supported ? There maye
be lots of "Access Denied" errors, or something undefined may happen.
I thought the "depends on IMX_SCU" was already added in Kconfig, that
means the module will be ONLY built with IMX_SCU enabled, and watchdog
will be always enabled if IMX_SCU is enabled. Is it safe enough?
No. The driver will be built with arm64:defconfig, meaning it will be built for
all arm64 systems using defconfig. Any such system will have the driver
installed as module, and nothing will prevent the user from running
modprobe. But even if it wasn't enabled with defconfig, we must not
instantiate the driver on an arbitrary system.
Is everyone on Cc: ok with this ? Is this how we handle
instantiations nowadays ? Or should the driver be instantiated from
imx_scu_probe() in drivers/firmware/imx/imx-scu.c ?
Sorry if the answer is obvious, but I am still struggling with "no
more instantiations through devicetree".
If the driver is auto-instantiated when the module is loaded, as
currently written, there needs to be some check if it is actually
supported, possibly in
imx_sc_wdt_init() or, if that is not possible, in the probe function.
I don't like that, but it would at least prevent the module from
being loaded when the hardware is not supported.
Other modules depend on the IMX_SCU IPC will have defer probe there to
make sure IMX_SCU driver is ready for IPC handle, since watchdog
driver ONLY uses ARM SMC but no IPC call, so I did NOT add the defer
probe handle here, so if adding it can answer your question/concern,
then I can add it, although I think the dependency in Kconfig should
be good here. Without SCU firmware ready, the SoC does NOT boot up A
core with ATF/Linux at all.
That has nothing to do with deferred probing. My concern is that the driver
will be instantiated just by loading it, no matter if the hardware supporting it
is present or not.
Having a devicetree node would have prevented that since it instantiates the
child drivers with devm_of_platform_populate() in imx_scu_probe(). The
equivalent is to register the platform device from imx_scu_probe().
That is the only means to prevent the driver from being instantiated where it
shouldn't. As mentioned before, I would have preferred the devicetree
method, but having it no longer available doesn't mean that we should add
risky code.
Maybe you can check in the init function if the 'fsl,imx-scu' node exists, and
instantiate the driver if it does. I don't like that either, but it would be
acceptable to me - and maybe better than instantiating it manually from
imx_scu_probe().
OK, I can add check of "fsl,imx-scu" in init function and ONLY instantiate the driver if
it exist, since there is no watchdog node in devicetree and "fsl,imx-scu" should be treated
as watchdog node, so adding such check also makes some sense, putting it inside SCU
driver is NOT that good I think, do you agree?
I dislike both. Putting it into the SCU driver probe function would be more
appropriate. Putting it into the watchdog driver is worse since it should
really only be instantiated if (and after) the parent driver is instantiated,
and the parent driver should be listed as its parent. That will get lost
if you put the check into the watchdog driver.
I would suggest to submit the patch for the SCU driver first. If that gets
rejected by its maintainer, we'll have to do it in the watchdog driver.
Guenter