Skip to contentSkip to conversation

dt-bindings: firmware: qcom,scm: support indicating SDI default state

PATCH Applied Robert Marko 8 participants 17 messages Aug 16, 2023 – Sep 20, 2023

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH v3 1/4 dt-bindings: firmware: qcom,scm: support indicating SDI default state

IPQ5018 has SDI (Secure Debug Image) enabled by TZ by default, and that means that WDT being asserted or just trying to reboot will hang the board in the debug mode and only pulling the power and repowering will help. Some IPQ4019 boards like Google WiFI have it enabled as well.

So, lets add a boolean property to indicate that SDI is enabled by default and thus needs to be disabled by the kernel.

---
Changes in v3:
* Change the property so it indicates that SDI has been enabled by default

---
 Documentation/devicetree/bindings/firmware/qcom,scm.yaml | 8 ++++++++
 1 file changed, 8 insertions(+)
Documentation/devicetree/bindings/firmware/qcom,scm.yaml +8 -0
index 4233ea839bfc..590bbbd61de5 100644
@@ -89,6 +89,14 @@ properties:
       protocol to handle sleeping SCM calls.
     maxItems: 1
 
+  qcom,sdi-enabled:
+    description:
+      Indicates that the SDI (Secure Debug Image) has been enabled by TZ
+      by default and it needs to be disabled.
+      If not disabled WDT assertion or reboot will cause the board to hang
+      in the debug mode.
+    type: boolean
+
   qcom,dload-mode:
     $ref: /schemas/types.yaml#/definitions/phandle-array
     items:
Signature · 1 line
2.41.0
#1 lore raw

PATCH v3 2/4 firmware: qcom_scm: disable SDI if required

IPQ5018 has SDI (Secure Debug Image) enabled by TZ by default, and that means that WDT being asserted or just trying to reboot will hang the board in the debug mode and only pulling the power and repowering will help. Some IPQ4019 boards like Google WiFI have it enabled as well.

Luckily, SDI can be disabled via an SCM call.

So, lets use the boolean DT property to identify boards that have SDI enabled by default and use the SCM call to disable SDI during SCM probe. It is important to disable it as soon as possible as we might have a WDT assertion at any time which would then leave the board in debug mode, thus disabling it during SCM removal is not enough.

---
Changes in v3:
* Squashed ("firmware: qcom: scm: Add SDI disable support") and ("firmware: qcom_scm: disable SDI if required")

---
 drivers/firmware/qcom_scm.c | 29 +++++++++++++++++++++++++++++
 drivers/firmware/qcom_scm.h |  1 +
 2 files changed, 30 insertions(+)
drivers/firmware/qcom_scm.c +29 -0
index 06fe8aca870d..de9d1a11d097 100644
@@ -403,6 +403,29 @@ int qcom_scm_set_remote_state(u32 state, u32 id)
 }
 EXPORT_SYMBOL_GPL(qcom_scm_set_remote_state);
 
+static int qcom_scm_disable_sdi(void)
+{
+	int ret;
+	struct qcom_scm_desc desc = {
+		.svc = QCOM_SCM_SVC_BOOT,
+		.cmd = QCOM_SCM_BOOT_SDI_CONFIG,
+		.args[0] = 1, /* Disable watchdog debug */
+		.args[1] = 0, /* Disable SDI */
+		.arginfo = QCOM_SCM_ARGS(2),
+		.owner = ARM_SMCCC_OWNER_SIP,
+	};
+	struct qcom_scm_res res;
+
+	ret = qcom_scm_clk_enable();
+	if (ret)
+		return ret;
+	ret = qcom_scm_call(__scm->dev, &desc, &res);
+
+	qcom_scm_clk_disable();
+
+	return ret ? : res.result[0];
+}
+
 static int __qcom_scm_set_dload_mode(struct device *dev, bool enable)
 {
 	struct qcom_scm_desc desc = {
@@ -1468,6 +1491,12 @@ static int qcom_scm_probe(struct platform_device *pdev)
 	if (download_mode)
 		qcom_scm_set_download_mode(true);
 
+	/*
+	 * Disable SDI if indicated by DT that it is enabled by default.
+	 */
+	if (of_property_read_bool(pdev->dev.of_node, "qcom,sdi-enabled"))
+		qcom_scm_disable_sdi();
+
 	return 0;
 }
 
drivers/firmware/qcom_scm.h +1 -0
index e6e512bd57d1..7b68fa820495 100644
@@ -80,6 +80,7 @@ extern int scm_legacy_call(struct device *dev, const struct qcom_scm_desc *desc,
 #define QCOM_SCM_SVC_BOOT		0x01
 #define QCOM_SCM_BOOT_SET_ADDR		0x01
 #define QCOM_SCM_BOOT_TERMINATE_PC	0x02
+#define QCOM_SCM_BOOT_SDI_CONFIG	0x09
 #define QCOM_SCM_BOOT_SET_DLOAD_MODE	0x10
 #define QCOM_SCM_BOOT_SET_ADDR_MC	0x11
 #define QCOM_SCM_BOOT_SET_REMOTE_STATE	0x0a
Signature · 1 line
2.41.0
Reviewed-by on v3 2/4
#2 lore raw

Re: [PATCH v3 2/4] firmware: qcom_scm: disable SDI if required

Quoted 7 lines · Robert Marko

IPQ5018 has SDI (Secure Debug Image) enabled by TZ by default, and that means that WDT being asserted or just trying to reboot will hang the board in the debug mode and only pulling the power and repowering will help. Some IPQ4019 boards like Google WiFI have it enabled as well.

Luckily, SDI can be disabled via an SCM call.

Quoted 7 lines

So, lets use the boolean DT property to identify boards that have SDI enabled by default and use the SCM call to disable SDI during SCM probe. It is important to disable it as soon as possible as we might have a WDT assertion at any time which would then leave the board in debug mode, thus disabling it during SCM removal is not enough.

#3 lore raw

Re: [PATCH v3 2/4] firmware: qcom_scm: disable SDI if required

In reply to Robert Marko
Quoted 15 lines · Robert Marko

IPQ5018 has SDI (Secure Debug Image) enabled by TZ by default, and that means that WDT being asserted or just trying to reboot will hang the board in the debug mode and only pulling the power and repowering will help. Some IPQ4019 boards like Google WiFI have it enabled as well.

Luckily, SDI can be disabled via an SCM call.

So, lets use the boolean DT property to identify boards that have SDI enabled by default and use the SCM call to disable SDI during SCM probe. It is important to disable it as soon as possible as we might have a WDT assertion at any time which would then leave the board in debug mode, thus disabling it during SCM removal is not enough.

---

[...]

drivers/firmware/qcom_scm.c static int qcom_scm_probe(struct platform_device *pdev) · line 1498

Quoted 5 lines
drivers/firmware/qcom_scm.c +5 -0
+	/*
+	 * Disable SDI if indicated by DT that it is enabled by default.
+	 */
+	if (of_property_read_bool(pdev->dev.of_node, "qcom,sdi-enabled"))
+		qcom_scm_disable_sdi();

Should we care about the return value?

Konrad

#4 lore raw

Re: [PATCH v3 2/4] firmware: qcom_scm: disable SDI if required

Quoted 20 lines · Konrad Dybcio
Quoted 15 lines · Robert Marko

IPQ5018 has SDI (Secure Debug Image) enabled by TZ by default, and that means that WDT being asserted or just trying to reboot will hang the board in the debug mode and only pulling the power and repowering will help. Some IPQ4019 boards like Google WiFI have it enabled as well.

Luckily, SDI can be disabled via an SCM call.

So, lets use the boolean DT property to identify boards that have SDI enabled by default and use the SCM call to disable SDI during SCM probe. It is important to disable it as soon as possible as we might have a WDT assertion at any time which would then leave the board in debug mode, thus disabling it during SCM removal is not enough.

---

[...]

Quoted 6 lines

drivers/firmware/qcom_scm.c static int qcom_scm_probe(struct platform_device *pdev) · line 1498

Quoted 5 lines
drivers/firmware/qcom_scm.c +5 -0
+     /*
+      * Disable SDI if indicated by DT that it is enabled by default.
+      */
+     if (of_property_read_bool(pdev->dev.of_node, "qcom,sdi-enabled"))
+             qcom_scm_disable_sdi();

Should we care about the return value?

I dont really see a point, as we cant really do anything about it if it fails. Also, the SDI SCM call seems to have weird return codes, for example, it works but it returns 2.

Maybe somebody from QCA can shine some light on that?

Regards,
Robert

Quoted 2 lines

Konrad

#5 lore raw

Re: [PATCH v3 2/4] firmware: qcom_scm: disable SDI if required

In reply to Robert Marko
Quoted 38 lines · Robert Marko

IPQ5018 has SDI (Secure Debug Image) enabled by TZ by default, and that means that WDT being asserted or just trying to reboot will hang the board in the debug mode and only pulling the power and repowering will help. Some IPQ4019 boards like Google WiFI have it enabled as well.

Luckily, SDI can be disabled via an SCM call.

So, lets use the boolean DT property to identify boards that have SDI enabled by default and use the SCM call to disable SDI during SCM probe. It is important to disable it as soon as possible as we might have a WDT assertion at any time which would then leave the board in debug mode, thus disabling it during SCM removal is not enough.

---
Changes in v3:
* Squashed ("firmware: qcom: scm: Add SDI disable support") and ("firmware: qcom_scm: disable SDI if required")

---
  drivers/firmware/qcom_scm.c | 29 +++++++++++++++++++++++++++++
  drivers/firmware/qcom_scm.h |  1 +
  2 files changed, 30 insertions(+)
drivers/firmware/qcom_scm.c +7 -0
index 06fe8aca870d..de9d1a11d097 100644
@@ -403,6 +403,29 @@ int qcom_scm_set_remote_state(u32 state, u32 id)
  }
  EXPORT_SYMBOL_GPL(qcom_scm_set_remote_state);
 
+static int qcom_scm_disable_sdi(void)
+{
+	int ret;
+	struct qcom_scm_desc desc = {
+		.svc = QCOM_SCM_SVC_BOOT,
+		.cmd = QCOM_SCM_BOOT_SDI_CONFIG,
+		.args[0] = 1, /* Disable watchdog debug */

drivers/firmware/qcom_scm.c int qcom_scm_set_remote_state(u32 state, u32 id) · line 422

Quoted 10 lines
drivers/firmware/qcom_scm.c +10 -0
+		.args[1] = 0, /* Disable SDI */
+		.arginfo = QCOM_SCM_ARGS(2),
+		.owner = ARM_SMCCC_OWNER_SIP,
+	};
+	struct qcom_scm_res res;
+
+	ret = qcom_scm_clk_enable();
+	if (ret)
+		return ret;
+	ret = qcom_scm_call(__scm->dev, &desc, &res);

Would you not be wanting this call to be atomic ?

Quoted 10 lines
drivers/firmware/qcom_scm.c +6 -0
+
+	qcom_scm_clk_disable();
+
+	return ret ? : res.result[0];
+}
+
  static int __qcom_scm_set_dload_mode(struct device *dev, bool enable)
  {
  	struct qcom_scm_desc desc = {

drivers/firmware/qcom_scm.c static int qcom_scm_probe(struct platform_device *pdev) · line 1498

Quoted 8 lines
drivers/firmware/qcom_scm.c +5 -0
@@ -1468,6 +1491,12 @@ static int qcom_scm_probe(struct platform_device *pdev)
  	if (download_mode)
  		qcom_scm_set_download_mode(true);
 
+	/*
+	 * Disable SDI if indicated by DT that it is enabled by default.
+	 */
+	if (of_property_read_bool(pdev->dev.of_node, "qcom,sdi-enabled"))
+		qcom_scm_disable_sdi();

Why don't we do this call in qcom_scm_shutdown() also does it not conflict with above download_mode we have enabled download mode but disabling SDI means (hard reset) and will not be collecting crash dump?

-Mukesh

Quoted 16 lines
drivers/firmware/qcom_scm.c +1 -0
+
  	return 0;
  }
 
drivers/firmware/qcom_scm.h +1 -0
index e6e512bd57d1..7b68fa820495 100644
@@ -80,6 +80,7 @@ extern int scm_legacy_call(struct device *dev, const struct qcom_scm_desc *desc,
  #define QCOM_SCM_SVC_BOOT		0x01
  #define QCOM_SCM_BOOT_SET_ADDR		0x01
  #define QCOM_SCM_BOOT_TERMINATE_PC	0x02
+#define QCOM_SCM_BOOT_SDI_CONFIG	0x09
  #define QCOM_SCM_BOOT_SET_DLOAD_MODE	0x10
  #define QCOM_SCM_BOOT_SET_ADDR_MC	0x11
  #define QCOM_SCM_BOOT_SET_REMOTE_STATE	0x0a
#6 lore raw

Re: [PATCH v3 2/4] firmware: qcom_scm: disable SDI if required

Quoted 46 lines · Mukesh Ojha
Quoted 42 lines · Robert Marko

IPQ5018 has SDI (Secure Debug Image) enabled by TZ by default, and that means that WDT being asserted or just trying to reboot will hang the board in the debug mode and only pulling the power and repowering will help. Some IPQ4019 boards like Google WiFI have it enabled as well.

Luckily, SDI can be disabled via an SCM call.

So, lets use the boolean DT property to identify boards that have SDI enabled by default and use the SCM call to disable SDI during SCM probe. It is important to disable it as soon as possible as we might have a WDT assertion at any time which would then leave the board in debug mode, thus disabling it during SCM removal is not enough.

---
Changes in v3:
* Squashed ("firmware: qcom: scm: Add SDI disable support") and ("firmware: qcom_scm: disable SDI if required")

---
  drivers/firmware/qcom_scm.c | 29 +++++++++++++++++++++++++++++
  drivers/firmware/qcom_scm.h |  1 +
  2 files changed, 30 insertions(+)
drivers/firmware/qcom_scm.c +11 -0
index 06fe8aca870d..de9d1a11d097 100644
@@ -403,6 +403,29 @@ int qcom_scm_set_remote_state(u32 state, u32 id)
  }
  EXPORT_SYMBOL_GPL(qcom_scm_set_remote_state);
 
+static int qcom_scm_disable_sdi(void)
+{
+     int ret;
+     struct qcom_scm_desc desc = {
+             .svc = QCOM_SCM_SVC_BOOT,
+             .cmd = QCOM_SCM_BOOT_SDI_CONFIG,
+             .args[0] = 1, /* Disable watchdog debug */
+             .args[1] = 0, /* Disable SDI */
+             .arginfo = QCOM_SCM_ARGS(2),
+             .owner = ARM_SMCCC_OWNER_SIP,
+     };
Quoted 8 lines

drivers/firmware/qcom_scm.c int qcom_scm_set_remote_state(u32 state, u32 id) · line 422

Quoted 6 lines
drivers/firmware/qcom_scm.c +6 -0
+     struct qcom_scm_res res;
+
+     ret = qcom_scm_clk_enable();
+     if (ret)
+             return ret;
+     ret = qcom_scm_call(__scm->dev, &desc, &res);

Would you not be wanting this call to be atomic ?

This is implemented based off the downstream 5.4 kernel as I dont have the SCM docs
so I dont know if its even supported in the atomic version.

Quoted 14 lines
Quoted 14 lines
drivers/firmware/qcom_scm.c +6 -0
+
+     qcom_scm_clk_disable();
+
+     return ret ? : res.result[0];
+}
+
  static int __qcom_scm_set_dload_mode(struct device *dev, bool enable)
  {
      struct qcom_scm_desc desc = {
@@ -1468,6 +1491,12 @@ static int qcom_scm_probe(struct platform_device *pdev)
      if (download_mode)
              qcom_scm_set_download_mode(true);
 
Quoted 11 lines

drivers/firmware/qcom_scm.c static int qcom_scm_probe(struct platform_device *pdev) · line 1498

Quoted 5 lines
drivers/firmware/qcom_scm.c +5 -0
+     /*
+      * Disable SDI if indicated by DT that it is enabled by default.
+      */
+     if (of_property_read_bool(pdev->dev.of_node, "qcom,sdi-enabled"))
+             qcom_scm_disable_sdi();

Why don't we do this call in qcom_scm_shutdown() also does it not conflict with above download_mode we have enabled download mode but disabling SDI means (hard reset) and will not be collecting crash dump?

Because doing it in SCM removal is too late, what if we have a WDT assertion and not a
regular reboot?
It would mean that the board will get stuck in the debug mode which is not useful for users and
requires the power to be pulled in order to boot normally again.

I am not sure about the download mode, this is where insight from QCA really help as I am
doing this with very limited docs.

Regards,
Robert

Quoted 19 lines

-Mukesh

Quoted 16 lines
drivers/firmware/qcom_scm.c +1 -0
+
      return 0;
  }
 
drivers/firmware/qcom_scm.h +1 -0
index e6e512bd57d1..7b68fa820495 100644
@@ -80,6 +80,7 @@ extern int scm_legacy_call(struct device *dev, const struct qcom_scm_desc *desc,
  #define QCOM_SCM_SVC_BOOT           0x01
  #define QCOM_SCM_BOOT_SET_ADDR              0x01
  #define QCOM_SCM_BOOT_TERMINATE_PC  0x02
+#define QCOM_SCM_BOOT_SDI_CONFIG     0x09
  #define QCOM_SCM_BOOT_SET_DLOAD_MODE        0x10
  #define QCOM_SCM_BOOT_SET_ADDR_MC   0x11
  #define QCOM_SCM_BOOT_SET_REMOTE_STATE      0x0a
#7 lore raw

Re: [PATCH v3 2/4] firmware: qcom_scm: disable SDI if required

Quoted 9 lines · Robert Marko
Quoted 4 lines · Mukesh Ojha
Quoted 1 line · Robert Marko

+ ret = qcom_scm_call(__scm->dev, &desc, &res);

Would you not be wanting this call to be atomic ?

This is implemented based off the downstream 5.4 kernel as I dont have the SCM docs
so I dont know if its even supported in the atomic version.

More than that, it's known *not* to be properly supported on one such applicable device:

Subject: [RFC] qcom_scm: IPQ4019 firmware does not support atomic API?
https://lore.kernel.org/linux-arm-kernel/20200913201608.GA3162100@bDebian/

I still haven't gotten a solution to *that* problem upstream, but it'd be nice not to make it worse.

Brian

#8 lore raw

Re: [PATCH v3 2/4] firmware: qcom_scm: disable SDI if required

In reply to Robert Marko
Quoted 56 lines · Robert Marko
Quoted 46 lines · Mukesh Ojha
Quoted 42 lines · Robert Marko

IPQ5018 has SDI (Secure Debug Image) enabled by TZ by default, and that means that WDT being asserted or just trying to reboot will hang the board in the debug mode and only pulling the power and repowering will help. Some IPQ4019 boards like Google WiFI have it enabled as well.

Luckily, SDI can be disabled via an SCM call.

So, lets use the boolean DT property to identify boards that have SDI enabled by default and use the SCM call to disable SDI during SCM probe. It is important to disable it as soon as possible as we might have a WDT assertion at any time which would then leave the board in debug mode, thus disabling it during SCM removal is not enough.

---
Changes in v3:
* Squashed ("firmware: qcom: scm: Add SDI disable support") and ("firmware: qcom_scm: disable SDI if required")

---
   drivers/firmware/qcom_scm.c | 29 +++++++++++++++++++++++++++++
   drivers/firmware/qcom_scm.h |  1 +
   2 files changed, 30 insertions(+)
drivers/firmware/qcom_scm.c +11 -0
index 06fe8aca870d..de9d1a11d097 100644
@@ -403,6 +403,29 @@ int qcom_scm_set_remote_state(u32 state, u32 id)
   }
   EXPORT_SYMBOL_GPL(qcom_scm_set_remote_state);
 
+static int qcom_scm_disable_sdi(void)
+{
+     int ret;
+     struct qcom_scm_desc desc = {
+             .svc = QCOM_SCM_SVC_BOOT,
+             .cmd = QCOM_SCM_BOOT_SDI_CONFIG,
+             .args[0] = 1, /* Disable watchdog debug */
+             .args[1] = 0, /* Disable SDI */
+             .arginfo = QCOM_SCM_ARGS(2),
+             .owner = ARM_SMCCC_OWNER_SIP,
+     };
Quoted 8 lines

drivers/firmware/qcom_scm.c int qcom_scm_set_remote_state(u32 state, u32 id) · line 422

Quoted 6 lines
drivers/firmware/qcom_scm.c +6 -0
+     struct qcom_scm_res res;
+
+     ret = qcom_scm_clk_enable();
+     if (ret)
+             return ret;
+     ret = qcom_scm_call(__scm->dev, &desc, &res);

Would you not be wanting this call to be atomic ?

Quoted 3 lines

This is implemented based off the downstream 5.4 kernel as I dont have the SCM docs
so I dont know if its even supported in the atomic version.

Ok,.

Well, Kernel version does not guarantees us whether certain things are supported or not in the firmware and it is not bound to any particular firmware version;

So, whatever firmware version it is running with, we should try to support.

Should we implement certain kind of call, if fastcall(atomic) is supported go-ahead otherwise fallback to slowcalls (interruptible) calls, but this is completely out of the context of this patch.

Quoted 26 lines
Quoted 14 lines
Quoted 14 lines
drivers/firmware/qcom_scm.c +6 -0
+
+     qcom_scm_clk_disable();
+
+     return ret ? : res.result[0];
+}
+
   static int __qcom_scm_set_dload_mode(struct device *dev, bool enable)
   {
       struct qcom_scm_desc desc = {
@@ -1468,6 +1491,12 @@ static int qcom_scm_probe(struct platform_device *pdev)
       if (download_mode)
               qcom_scm_set_download_mode(true);
 
Quoted 11 lines

drivers/firmware/qcom_scm.c static int qcom_scm_probe(struct platform_device *pdev) · line 1498

Quoted 5 lines
drivers/firmware/qcom_scm.c +5 -0
+     /*
+      * Disable SDI if indicated by DT that it is enabled by default.
+      */
+     if (of_property_read_bool(pdev->dev.of_node, "qcom,sdi-enabled"))
+             qcom_scm_disable_sdi();

Why don't we do this call in qcom_scm_shutdown() also does it not conflict with above download_mode we have enabled download mode but disabling SDI means (hard reset) and will not be collecting crash dump?

Quoted 6 lines

Because doing it in SCM removal is too late, what if we have a WDT assertion and not a
regular reboot?
It would mean that the board will get stuck in the debug mode which is not useful for users and
requires the power to be pulled in order to boot normally again.

Agree.

Just a wild guess..

Can we check if this call __qcom_scm_is_call_available() helps to determine, if the certain soc has this SCM calls supported and if it is there it can be disabled.

__qcom_scm_is_call_available(__scm->dev, QCOM_SCM_SVC_BOOT, QCOM_SCM_BOOT_SDI_CONFIG)

Quoted 4 lines

I am not sure about the download mode, this is where insight from QCA really help as I am
doing this with very limited docs.

Download mode would not be reflected unless it is debug board, whatever you write will not be allowed if it is a secure device.

-Mukesh

Quoted 22 lines

Regards,
Robert

Quoted 19 lines

-Mukesh

Quoted 16 lines
drivers/firmware/qcom_scm.c +1 -0
+
       return 0;
   }
 
drivers/firmware/qcom_scm.h +1 -0
index e6e512bd57d1..7b68fa820495 100644
@@ -80,6 +80,7 @@ extern int scm_legacy_call(struct device *dev, const struct qcom_scm_desc *desc,
   #define QCOM_SCM_SVC_BOOT           0x01
   #define QCOM_SCM_BOOT_SET_ADDR              0x01
   #define QCOM_SCM_BOOT_TERMINATE_PC  0x02
+#define QCOM_SCM_BOOT_SDI_CONFIG     0x09
   #define QCOM_SCM_BOOT_SET_DLOAD_MODE        0x10
   #define QCOM_SCM_BOOT_SET_ADDR_MC   0x11
   #define QCOM_SCM_BOOT_SET_REMOTE_STATE      0x0a
#9 lore raw

Re: [PATCH v3 2/4] firmware: qcom_scm: disable SDI if required

<snip ..>

Quoted 28 lines
Quoted 16 lines
Quoted 16 lines
Quoted 15 lines

+     int ret;
+     struct qcom_scm_desc desc = {
+             .svc = QCOM_SCM_SVC_BOOT,
+             .cmd = QCOM_SCM_BOOT_SDI_CONFIG, +             .args[0] = 1, /* Disable watchdog debug */ +             .args[1] = 0, /* Disable SDI */
+             .arginfo = QCOM_SCM_ARGS(2),
+             .owner = ARM_SMCCC_OWNER_SIP,
+     };
+     struct qcom_scm_res res;
+
+     ret = qcom_scm_clk_enable();
+     if (ret)
+             return ret;
+     ret = qcom_scm_call(__scm->dev, &desc, &res);

Quoted 5 lines
Quoted 1 line

Would you not be wanting this call to be atomic ?

This is implemented based off the downstream 5.4 kernel as I dont have the SCM docs
so I dont know if its even supported in the atomic version.

Ok,.

Well, Kernel version does not guarantees us whether certain things are supported or not in the firmware and it is not bound to any particular firmware version;

Quoted 6 lines

So, whatever firmware version it is running with, we should try to support.

Should we implement certain kind of call, if fastcall(atomic) is supported go-ahead otherwise fallback to slowcalls (interruptible) calls, but this is completely out of the context of this patch.

I replied on older thread, was not in CC here, just saw this.

Agree, atomic api is out of this context and we could take it up separately.

Quoted 27 lines
Quoted 27 lines
Quoted 26 lines
Quoted 11 lines
drivers/firmware/qcom_scm.c +6 -0
+
+     qcom_scm_clk_disable();
+
+     return ret ? : res.result[0];
+}
+

   static int __qcom_scm_set_dload_mode(struct device *dev, bool enable) {
       struct qcom_scm_desc desc = {

drivers/firmware/qcom_scm.c static int qcom_scm_probe(struct · line 1499

Quoted 9 lines
drivers/firmware/qcom_scm.c +5 -0
@@ -1468,6 +1491,12 @@ static int qcom_scm_probe(struct 
 platform_device *pdev)
        if (download_mode)
                qcom_scm_set_download_mode(true);
 
+     /*
+      * Disable SDI if indicated by DT that it is enabled by default.
+      */
+     if (of_property_read_bool(pdev->dev.of_node, "qcom,sdi-enabled"))
+             qcom_scm_disable_sdi();

Why don't we do this call in qcom_scm_shutdown() also does it not conflict with above download_mode we have enabled download mode but disabling SDI means (hard reset) and will not be collecting crash dump?

Quoted 8 lines
Quoted 6 lines

Because doing it in SCM removal is too late, what if we have a WDT assertion and not a
regular reboot?
It would mean that the board will get stuck in the debug mode which is not useful for users and
requires the power to be pulled in order to boot normally again.

Agree.

For IPQ chipsets, SDI bit is used like below,

For abnormal resets (like WDT), should be set '1' for valid dump collection.

For reboot, should be cleared to '0' to avoid dump collection which is not required in this case.

For HLOS panic, is a don't care, dumps always get collected and firmware takes care of clearing the SDI bit.

Mukesh, Can you confirm if its same for msm also ?

Quoted 9 lines

Just a wild guess..

Can we check if this call __qcom_scm_is_call_available() helps to determine, if the certain soc has this SCM calls supported and if it is there it can be disabled.

__qcom_scm_is_call_available(__scm->dev, QCOM_SCM_SVC_BOOT, QCOM_SCM_BOOT_SDI_CONFIG)

Yes, as i mentioned in other thread, checking using qcom_scm_is_call_available is better. That said, would require testing on all IPQ/MSM socs to confirm if firmware supports it.

Quoted 9 lines
Quoted 4 lines

I am not sure about the download mode, this is where insight from QCA really help as I am
doing this with very limited docs.

Download mode would not be reflected unless it is debug board, whatever you write will not be allowed if it is a secure device.

Yes, 'download mode' bit is similar, but that is used by the firmware to determining whether to collect dumps on non-secure boards. Specifically, 'SDI bit' on some socs is used by firmware to determine if boot is happening from a 'abnormal crash', hence put DDR to self-refresh etc for valid dumps.

Regards,
Sricharan

#10 lore raw

Re: [PATCH v3 2/4] firmware: qcom_scm: disable SDI if required

Quoted 85 lines · Sricharan Ramabadhran

<snip ..>

Quoted 28 lines
Quoted 16 lines
Quoted 16 lines
Quoted 15 lines

+     int ret;
+     struct qcom_scm_desc desc = {
+             .svc = QCOM_SCM_SVC_BOOT,
+             .cmd = QCOM_SCM_BOOT_SDI_CONFIG, +             .args[0] = 1, /* Disable watchdog debug */ +             .args[1] = 0, /* Disable SDI */
+             .arginfo = QCOM_SCM_ARGS(2),
+             .owner = ARM_SMCCC_OWNER_SIP,
+     };
+     struct qcom_scm_res res;
+
+     ret = qcom_scm_clk_enable();
+     if (ret)
+             return ret;
+     ret = qcom_scm_call(__scm->dev, &desc, &res);

Quoted 5 lines
Quoted 1 line

Would you not be wanting this call to be atomic ?

This is implemented based off the downstream 5.4 kernel as I dont have the SCM docs
so I dont know if its even supported in the atomic version.

Ok,.

Well, Kernel version does not guarantees us whether certain things are supported or not in the firmware and it is not bound to any particular firmware version;

Quoted 6 lines

So, whatever firmware version it is running with, we should try to support.

Should we implement certain kind of call, if fastcall(atomic) is supported go-ahead otherwise fallback to slowcalls (interruptible) calls, but this is completely out of the context of this patch.

 I replied on older thread, was not in CC here, just saw this.

 Agree, atomic api is out of this context and we could take it up separately.

Quoted 28 lines
Quoted 28 lines
Quoted 27 lines
Quoted 11 lines
drivers/firmware/qcom_scm.c +6 -0
+
+     qcom_scm_clk_disable();
+
+     return ret ? : res.result[0];
+}
+

   static int __qcom_scm_set_dload_mode(struct device *dev, bool enable) {
       struct qcom_scm_desc desc = {

drivers/firmware/qcom_scm.c static int qcom_scm_probe(struct · line 1500

Quoted 10 lines
drivers/firmware/qcom_scm.c +5 -0
@@ -1468,6 +1491,12 @@ static int qcom_scm_probe(struct 
 platform_device *pdev)
        if (download_mode)
                qcom_scm_set_download_mode(true);
 
+     /*
+      * Disable SDI if indicated by DT that it is enabled by default.
+      */
+     if (of_property_read_bool(pdev->dev.of_node, 
 "qcom,sdi-enabled"))
+             qcom_scm_disable_sdi();

Why don't we do this call in qcom_scm_shutdown() also does it not conflict with above download_mode we have enabled download mode but disabling SDI means (hard reset) and will not be collecting crash dump?

Quoted 8 lines
Quoted 6 lines

Because doing it in SCM removal is too late, what if we have a WDT assertion and not a
regular reboot?
It would mean that the board will get stuck in the debug mode which is not useful for users and
requires the power to be pulled in order to boot normally again.

Agree.

 For IPQ chipsets, SDI bit is used like below,

   For abnormal resets (like WDT), should be set '1' for valid dump collection.

Quoted 7 lines

   For reboot, should be cleared to '0' to avoid dump collection which is not required in this case.

   For HLOS panic, is a don't care, dumps always get collected and firmware takes care of clearing the SDI bit.

   Mukesh,  Can you confirm if its same for msm also ?

Yes, it is same in MSM as well.

-Mukesh

Quoted 32 lines
Quoted 9 lines

Just a wild guess..

Can we check if this call __qcom_scm_is_call_available() helps to determine, if the certain soc has this SCM calls supported and if it is there it can be disabled.

__qcom_scm_is_call_available(__scm->dev, QCOM_SCM_SVC_BOOT, QCOM_SCM_BOOT_SDI_CONFIG)

 Yes, as i mentioned in other thread, checking using qcom_scm_is_call_available is better. That said, would require testing on all IPQ/MSM socs to confirm if firmware supports it.

Quoted 9 lines
Quoted 4 lines

I am not sure about the download mode, this is where insight from QCA really help as I am
doing this with very limited docs.

Download mode would not be reflected unless it is debug board, whatever you write will not be allowed if it is a secure device.

  Yes, 'download mode' bit is similar, but that is used by the firmware to determining whether to collect dumps on non-secure boards. Specifically, 'SDI bit' on some socs is used by firmware to determine if boot is happening from a 'abnormal crash', hence put DDR to self-refresh etc for valid dumps.

Regards,
 Sricharan

#11 lore raw

PATCH v3 3/4 dt-bindings: firmware: qcom,scm: document IPQ5018 compatible

In reply to Robert Marko

It seems that IPQ5018 compatible was never documented in the bindings.

---
 Documentation/devicetree/bindings/firmware/qcom,scm.yaml | 1 +
 1 file changed, 1 insertion(+)
Documentation/devicetree/bindings/firmware/qcom,scm.yaml +1 -0
index 590bbbd61de5..60c66cbcfd7b 100644
@@ -24,6 +24,7 @@ properties:
           - qcom,scm-apq8064
           - qcom,scm-apq8084
           - qcom,scm-ipq4019
+          - qcom,scm-ipq5018
           - qcom,scm-ipq5332
           - qcom,scm-ipq6018
           - qcom,scm-ipq806x
Signature · 1 line
2.41.0
#12 lore raw

PATCH v3 4/4 arm64: dts: qcom: ipq5018: indicate that SDI should be disabled

In reply to Robert Marko

Now that SCM has support for indicating that SDI has been enabled by default, lets set the property so SCM disables it during probing.

---
Changes in v3:
* Use the new boolean property

---
 arch/arm64/boot/dts/qcom/ipq5018.dtsi | 1 +
 1 file changed, 1 insertion(+)
arch/arm64/boot/dts/qcom/ipq5018.dtsi +1 -0
index 9f13d2dcdfd5..bd0198aaf8d0 100644
@@ -57,6 +57,7 @@ L2_0: l2-cache {
 	firmware {
 		scm {
 			compatible = "qcom,scm-ipq5018", "qcom,scm";
+			qcom,sdi-enabled;
 		};
 	};
 
Signature · 1 line
2.41.0
Reviewed-by on v3 1/4
#13 lore raw
In reply to Robert Marko
Quoted 9 lines · Robert Marko

IPQ5018 has SDI (Secure Debug Image) enabled by TZ by default, and that means that WDT being asserted or just trying to reboot will hang the board in the debug mode and only pulling the power and repowering will help. Some IPQ4019 boards like Google WiFI have it enabled as well.

So, lets add a boolean property to indicate that SDI is enabled by default and thus needs to be disabled by the kernel.

The series looks good to me. Thanks for doing this!

Reviewed-by on v3 1/4
#14 lore raw
In reply to Robert Marko
Quoted 9 lines · Robert Marko

IPQ5018 has SDI (Secure Debug Image) enabled by TZ by default, and that means that WDT being asserted or just trying to reboot will hang the board in the debug mode and only pulling the power and repowering will help. Some IPQ4019 boards like Google WiFI have it enabled as well.

So, lets add a boolean property to indicate that SDI is enabled by default and thus needs to be disabled by the kernel.

Best regards,
Krzysztof

Acked-by on v3 1/4
#15 lore raw
In reply to Robert Marko
Quoted 9 lines · Robert Marko

IPQ5018 has SDI (Secure Debug Image) enabled by TZ by default, and that means that WDT being asserted or just trying to reboot will hang the board in the debug mode and only pulling the power and repowering will help. Some IPQ4019 boards like Google WiFI have it enabled as well.

So, lets add a boolean property to indicate that SDI is enabled by default and thus needs to be disabled by the kernel.

-Mukesh

Quoted 26 lines

---
Changes in v3:
* Change the property so it indicates that SDI has been enabled by default

---
  Documentation/devicetree/bindings/firmware/qcom,scm.yaml | 8 ++++++++
  1 file changed, 8 insertions(+)
Documentation/devicetree/bindings/firmware/qcom,scm.yaml +8 -0
index 4233ea839bfc..590bbbd61de5 100644
@@ -89,6 +89,14 @@ properties:
        protocol to handle sleeping SCM calls.
      maxItems: 1
 
+  qcom,sdi-enabled:
+    description:
+      Indicates that the SDI (Secure Debug Image) has been enabled by TZ
+      by default and it needs to be disabled.
+      If not disabled WDT assertion or reboot will cause the board to hang
+      in the debug mode.
+    type: boolean
+
    qcom,dload-mode:
      $ref: /schemas/types.yaml#/definitions/phandle-array
      items:
maintainer Applied
#16 lore raw

Re: (subset) [PATCH v3 1/4] dt-bindings: firmware: qcom,scm: support indicating SDI default state

In reply to Robert Marko
Quoted 9 lines · Robert Marko

IPQ5018 has SDI (Secure Debug Image) enabled by TZ by default, and that means that WDT being asserted or just trying to reboot will hang the board in the debug mode and only pulling the power and repowering will help. Some IPQ4019 boards like Google WiFI have it enabled as well.

So, lets add a boolean property to indicate that SDI is enabled by default and thus needs to be disabled by the kernel.

[...]

Applied, thanks!

[4/4] arm64: dts: qcom: ipq5018: indicate that SDI should be disabled

      commit: 79796e87215db9587d6c66ec6f6781e091bc6464

Best regards,

Signature · 1 line
Bjorn Andersson <[email protected]>