From: Peng Fan <[email protected]>
V6:
Two changes are made:
1.
In V5, patch 3 has a wrong logic in imx_rproc_put_scu, see [1], however
patch 4 fixed the wrong logic, as below:
- if (!imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id))
- return;
+ if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id))
+ return imx_rproc_detach_pd(rproc);
In V6, patch 3 fixed the wrong logic, drop the fix in patch 4. And patch 4
changed as below, because put scu not has return value.
imx_rproc_detach_pd(rproc);
return;
2. Include patch 7/7 to enable attach recovery, because the attach recovery
feature has been supported in remoteproc core.
[1] https://lore.kernel.org/linux-remoteproc/DU0PR04MB941763C3EBF8AE9A9CD69445884C9@DU0PR04MB9417.eurprd04.prod.outlook.com/T/#m87e2864260d0a1c431b577068ba79e1d64c595f1
V5:
The patchset could not apply on latest remoteproc for-next branch,
so rebased. Only patch 6 has a minor conflict, other patches are not
changed.
V4:
https://lore.kernel.org/all/[email protected]/
Add R-b from DT maintainer
Fix probe failure and driver remove path in patch 3, 4
Add comments about i.MX8QM entries filter in patch 5
V3:
Drop the dependency in V2.
Tested on i.MX8QM/8MM
Use 'fsl,resource-id' and 'fsl,entry-address' Per dt maintainer
Drop 'reg' property Per remoteproc maintainer
Drop mcore self recovery, until we land in common framework support.
https://patchwork.kernel.org/project/linux-remoteproc/cover/[email protected]/
V2:
https://patchwork.kernel.org/project/linux-remoteproc/cover/[email protected]/
Depends on https://patchwork.kernel.org/project/linux-remoteproc/list/?series=621311
Tested on i.MX8QXP/QM/8MP
Addressed Mathieu's comments
Drop V1 patch 5/9, patch 3/9 is replaced with upper dependency patchset
Move V1 patch 4/9 out to https://patchwork.kernel.org/project/linux-remoteproc/patch/[email protected]/
Update commit log
Drop magic number to get entry address from device tree in patch 4/6
The V1 patchset:
https://patchwork.kernel.org/project/linux-remoteproc/patch/[email protected]/
Peng Fan (7):
dt-bindings: remoteproc: imx_rproc: support i.MX8QXP
dt-bindings: remoteproc: imx_rproc: support i.MX8QM
remoteproc: imx_rproc: support attaching to i.MX8QXP M4
remoteproc: imx_rproc: support kicking Mcore from Linux for i.MX8QXP
remoteproc: imx_rproc: support i.MX8QM
remoteproc: imx_rproc: request mbox channel later
remoteproc: imx_rproc: Enable attach recovery for i.MX8QM/QXP
.../bindings/remoteproc/fsl,imx-rproc.yaml | 16 +
drivers/remoteproc/imx_rproc.c | 277 +++++++++++++++++-
2 files changed, 287 insertions(+), 6 deletions(-)
--
2.37.1
From: Peng Fan <[email protected]>
i.MX8QM/QXP M4 could recover without help from Linux, so to support it:
- enable feature RPROC_FEAT_ATTACH_ON_RECOVERY
- add detach hook
- Since we have detach callback, we could move the free mbox operation
from partition reset notify to detach callback.
Signed-off-by: Peng Fan <[email protected]>
---
drivers/remoteproc/imx_rproc.c | 22 +++++++++++++++++++---
1 file changed, 19 insertions(+), 3 deletions(-)
diff --git a/drivers/remoteproc/imx_rproc.c b/drivers/remoteproc/imx_rproc.c
index bece44b46719..4057d6f33813 100644
--- a/drivers/remoteproc/imx_rproc.c
+++ b/drivers/remoteproc/imx_rproc.c
@@ -603,6 +603,22 @@ static int imx_rproc_attach(struct rproc *rproc)
return imx_rproc_xtr_mbox_init(rproc);
}
+static int imx_rproc_detach(struct rproc *rproc)
+{
+ struct imx_rproc *priv = rproc->priv;
+ const struct imx_rproc_dcfg *dcfg = priv->dcfg;
+
+ if (dcfg->method != IMX_RPROC_SCU_API)
+ return -EOPNOTSUPP;
+
+ if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id))
+ return -EOPNOTSUPP;
+
+ imx_rproc_free_mbox(rproc);
+
+ return 0;
+}
+
static struct resource_table *imx_rproc_get_loaded_rsc_table(struct rproc *rproc, size_t *table_sz)
{
struct imx_rproc *priv = rproc->priv;
@@ -618,6 +634,7 @@ static struct resource_table *imx_rproc_get_loaded_rsc_table(struct rproc *rproc
static const struct rproc_ops imx_rproc_ops = {
.prepare = imx_rproc_prepare,
.attach = imx_rproc_attach,
+ .detach = imx_rproc_detach,
.start = imx_rproc_start,
.stop = imx_rproc_stop,
.kick = imx_rproc_kick,
@@ -797,8 +814,6 @@ static int imx_rproc_partition_notify(struct notifier_block *nb,
if (!((event & BIT(priv->rproc_pt)) && (*(u8 *)group == IMX_SC_IRQ_GROUP_REBOOTED)))
return 0;
- imx_rproc_free_mbox(priv->rproc);
-
rproc_report_crash(priv->rproc, RPROC_WATCHDOG);
pr_info("Partition%d reset!\n", priv->rproc_pt);
@@ -916,7 +931,8 @@ static int imx_rproc_detect_mode(struct imx_rproc *priv)
}
priv->rproc->state = RPROC_DETACHED;
- priv->rproc->recovery_disabled = true;
+ priv->rproc->recovery_disabled = false;
+ rproc_set_feature(priv->rproc, RPROC_FEAT_ATTACH_ON_RECOVERY);
/* Get partition id and enable irq in SCFW */
ret = imx_sc_rm_get_resource_owner(priv->ipc_handle, priv->rsrc_id, &pt);
--
2.37.1
From: Peng Fan <[email protected]>
Add i.MX8QXP compatible
Add a new property fsl,resource-id for SoC which supports SCFW.
This property is used to check whether remote process is under control
of Linux or not.
Add fsl,entry-address to specify the entry address which used by SCFW
to kick M4.
To i.MX8QM/QXP, when M4 is in the same hardware partition with Cortex-A
cores, need power up M4 through SCFW, then M4 could start. So introduce
power-domains property
Reviewed-by: Rob Herring <[email protected]>
Signed-off-by: Peng Fan <[email protected]>
---
.../bindings/remoteproc/fsl,imx-rproc.yaml | 15 +++++++++++++++
1 file changed, 15 insertions(+)
diff --git a/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml b/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml
index 3a1f59ad79e2..70322e57b6ff 100644
--- a/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml
+++ b/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml
@@ -22,6 +22,7 @@ properties:
- fsl,imx8mn-cm7
- fsl,imx8mp-cm7
- fsl,imx8mq-cm4
+ - fsl,imx8qxp-cm4
- fsl,imx8ulp-cm33
- fsl,imx93-cm33
@@ -54,12 +55,26 @@ properties:
minItems: 1
maxItems: 32
+ power-domains:
+ maxItems: 8
+
fsl,auto-boot:
$ref: /schemas/types.yaml#/definitions/flag
description:
Indicate whether need to load the default firmware and start the remote
processor automatically.
+ fsl,entry-address:
+ $ref: /schemas/types.yaml#/definitions/uint32
+ description:
+ Specify CPU entry address for SCU enabled processor.
+
+ fsl,resource-id:
+ $ref: /schemas/types.yaml#/definitions/uint32
+ description:
+ This property is to specify the resource id of the remote processor in SoC
+ which supports SCFW
+
required:
- compatible
--
2.37.1
From: Peng Fan <[email protected]>
When M4 is in the same hardware partition with Cortex-A, it
could be start/stop by Linux.
Added power domain to make sure M4 could run, it requires several power
domains to work. Make clk always optional for i.MX8QXP, because
SCFW handles it when power up M4 core.
Signed-off-by: Peng Fan <[email protected]>
---
drivers/remoteproc/imx_rproc.c | 88 ++++++++++++++++++++++++++++++++--
1 file changed, 85 insertions(+), 3 deletions(-)
diff --git a/drivers/remoteproc/imx_rproc.c b/drivers/remoteproc/imx_rproc.c
index a56aecae00c6..7d2e11aed0aa 100644
--- a/drivers/remoteproc/imx_rproc.c
+++ b/drivers/remoteproc/imx_rproc.c
@@ -16,6 +16,7 @@
#include <linux/of_reserved_mem.h>
#include <linux/of_device.h>
#include <linux/platform_device.h>
+#include <linux/pm_domain.h>
#include <linux/regmap.h>
#include <linux/remoteproc.h>
#include <linux/workqueue.h>
@@ -79,6 +80,8 @@ struct imx_rproc_mem {
#define ATT_OWN BIT(1)
#define ATT_IOMEM BIT(2)
+static int imx_rproc_detach_pd(struct rproc *rproc);
+
struct imx_rproc {
struct device *dev;
struct regmap *regmap;
@@ -96,6 +99,10 @@ struct imx_rproc {
struct notifier_block rproc_nb;
u32 rproc_pt; /* partition id */
u32 rsrc_id; /* resource id */
+ u32 entry; /* cpu start address */
+ int num_pd;
+ struct device **pd_dev;
+ struct device_link **pd_dev_link;
};
static const struct imx_rproc_att imx_rproc_att_imx93[] = {
@@ -335,6 +342,9 @@ static int imx_rproc_start(struct rproc *rproc)
arm_smccc_smc(IMX_SIP_RPROC, IMX_SIP_RPROC_START, 0, 0, 0, 0, 0, 0, &res);
ret = res.a0;
break;
+ case IMX_RPROC_SCU_API:
+ ret = imx_sc_pm_cpu_start(priv->ipc_handle, priv->rsrc_id, true, priv->entry);
+ break;
default:
return -EOPNOTSUPP;
}
@@ -364,6 +374,9 @@ static int imx_rproc_stop(struct rproc *rproc)
if (res.a1)
dev_info(dev, "Not in wfi, force stopped\n");
break;
+ case IMX_RPROC_SCU_API:
+ ret = imx_sc_pm_cpu_start(priv->ipc_handle, priv->rsrc_id, false, priv->entry);
+ break;
default:
return -EOPNOTSUPP;
}
@@ -713,8 +726,10 @@ static void imx_rproc_put_scu(struct rproc *rproc)
if (dcfg->method != IMX_RPROC_SCU_API)
return;
- if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id))
+ if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id)) {
+ imx_rproc_detach_pd(rproc);
return;
+ }
imx_scu_irq_group_enable(IMX_SC_IRQ_GROUP_REBOOTED, BIT(priv->rproc_pt), false);
imx_scu_irq_unregister_notifier(&priv->rproc_nb);
@@ -736,6 +751,69 @@ static int imx_rproc_partition_notify(struct notifier_block *nb,
return 0;
}
+static int imx_rproc_attach_pd(struct imx_rproc *priv)
+{
+ struct device *dev = priv->dev;
+ int ret, i;
+
+ priv->num_pd = of_count_phandle_with_args(dev->of_node, "power-domains",
+ "#power-domain-cells");
+ if (priv->num_pd <= 1)
+ return 0;
+
+ priv->pd_dev = devm_kmalloc_array(dev, priv->num_pd, sizeof(*priv->pd_dev), GFP_KERNEL);
+ if (!priv->pd_dev)
+ return -ENOMEM;
+
+ priv->pd_dev_link = devm_kmalloc_array(dev, priv->num_pd, sizeof(*priv->pd_dev_link),
+ GFP_KERNEL);
+
+ if (!priv->pd_dev_link)
+ return -ENOMEM;
+
+ for (i = 0; i < priv->num_pd; i++) {
+ priv->pd_dev[i] = dev_pm_domain_attach_by_id(dev, i);
+ if (IS_ERR(priv->pd_dev[i])) {
+ ret = PTR_ERR(priv->pd_dev[i]);
+ goto detach_pd;
+ }
+
+ priv->pd_dev_link[i] = device_link_add(dev, priv->pd_dev[i], DL_FLAG_STATELESS |
+ DL_FLAG_PM_RUNTIME | DL_FLAG_RPM_ACTIVE);
+ if (!priv->pd_dev_link[i]) {
+ dev_pm_domain_detach(priv->pd_dev[i], false);
+ ret = -EINVAL;
+ goto detach_pd;
+ }
+ }
+
+ return 0;
+
+detach_pd:
+ while (--i >= 0) {
+ device_link_del(priv->pd_dev_link[i]);
+ dev_pm_domain_detach(priv->pd_dev[i], false);
+ }
+
+ return ret;
+}
+
+static int imx_rproc_detach_pd(struct rproc *rproc)
+{
+ struct imx_rproc *priv = rproc->priv;
+ int i;
+
+ if (priv->num_pd <= 1)
+ return 0;
+
+ for (i = 0; i < priv->num_pd; i++) {
+ device_link_del(priv->pd_dev_link[i]);
+ dev_pm_domain_detach(priv->pd_dev[i], false);
+ }
+
+ return 0;
+}
+
static int imx_rproc_detect_mode(struct imx_rproc *priv)
{
struct regmap_config config = { .name = "imx-rproc" };
@@ -770,8 +848,12 @@ static int imx_rproc_detect_mode(struct imx_rproc *priv)
* If Mcore resource is not owned by Acore partition, It is kicked by ROM,
* and Linux could only do IPC with Mcore and nothing else.
*/
- if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id))
- return 0;
+ if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id)) {
+ if (of_property_read_u32(dev->of_node, "fsl,entry-address", &priv->entry))
+ return -EINVAL;
+
+ return imx_rproc_attach_pd(priv);
+ }
priv->rproc->state = RPROC_DETACHED;
priv->rproc->recovery_disabled = true;
--
2.37.1
From: Peng Fan <[email protected]>
When M4 is kicked by SCFW, M4 runs in its own hardware partition, Linux
could only do IPC with M4, it could not start, stop, update image.
We disable recovery reboot when M4 is managed by SCFW, because
remoteproc core still not support M4 auto-recovery without loading
image.
Signed-off-by: Peng Fan <[email protected]>
---
drivers/remoteproc/imx_rproc.c | 108 ++++++++++++++++++++++++++++++++-
1 file changed, 107 insertions(+), 1 deletion(-)
diff --git a/drivers/remoteproc/imx_rproc.c b/drivers/remoteproc/imx_rproc.c
index 7cc4fd207e2d..a56aecae00c6 100644
--- a/drivers/remoteproc/imx_rproc.c
+++ b/drivers/remoteproc/imx_rproc.c
@@ -6,6 +6,7 @@
#include <linux/arm-smccc.h>
#include <linux/clk.h>
#include <linux/err.h>
+#include <linux/firmware/imx/sci.h>
#include <linux/interrupt.h>
#include <linux/kernel.h>
#include <linux/mailbox_client.h>
@@ -59,6 +60,8 @@
#define IMX_SIP_RPROC_STARTED 0x01
#define IMX_SIP_RPROC_STOP 0x02
+#define IMX_SC_IRQ_GROUP_REBOOTED 5
+
/**
* struct imx_rproc_mem - slim internal memory structure
* @cpu_addr: MPU virtual address of the memory region
@@ -89,6 +92,10 @@ struct imx_rproc {
struct work_struct rproc_work;
struct workqueue_struct *workqueue;
void __iomem *rsc_table;
+ struct imx_sc_ipc *ipc_handle;
+ struct notifier_block rproc_nb;
+ u32 rproc_pt; /* partition id */
+ u32 rsrc_id; /* resource id */
};
static const struct imx_rproc_att imx_rproc_att_imx93[] = {
@@ -117,6 +124,18 @@ static const struct imx_rproc_att imx_rproc_att_imx93[] = {
{ 0xD0000000, 0xa0000000, 0x10000000, 0 },
};
+static const struct imx_rproc_att imx_rproc_att_imx8qxp[] = {
+ { 0x08000000, 0x08000000, 0x10000000, 0 },
+ /* TCML/U */
+ { 0x1FFE0000, 0x34FE0000, 0x00040000, ATT_OWN | ATT_IOMEM },
+ /* OCRAM(Low 96KB) */
+ { 0x21000000, 0x00100000, 0x00018000, 0 },
+ /* OCRAM */
+ { 0x21100000, 0x00100000, 0x00040000, 0 },
+ /* DDR (Data) */
+ { 0x80000000, 0x80000000, 0x60000000, 0 },
+};
+
static const struct imx_rproc_att imx_rproc_att_imx8mn[] = {
/* dev addr , sys addr , size , flags */
/* ITCM */
@@ -255,6 +274,12 @@ static const struct imx_rproc_dcfg imx_rproc_cfg_imx8mq = {
.method = IMX_RPROC_MMIO,
};
+static const struct imx_rproc_dcfg imx_rproc_cfg_imx8qxp = {
+ .att = imx_rproc_att_imx8qxp,
+ .att_size = ARRAY_SIZE(imx_rproc_att_imx8qxp),
+ .method = IMX_RPROC_SCU_API,
+};
+
static const struct imx_rproc_dcfg imx_rproc_cfg_imx8ulp = {
.att = imx_rproc_att_imx8ulp,
.att_size = ARRAY_SIZE(imx_rproc_att_imx8ulp),
@@ -680,6 +705,37 @@ static void imx_rproc_free_mbox(struct rproc *rproc)
mbox_free_channel(priv->rx_ch);
}
+static void imx_rproc_put_scu(struct rproc *rproc)
+{
+ struct imx_rproc *priv = rproc->priv;
+ const struct imx_rproc_dcfg *dcfg = priv->dcfg;
+
+ if (dcfg->method != IMX_RPROC_SCU_API)
+ return;
+
+ if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id))
+ return;
+
+ imx_scu_irq_group_enable(IMX_SC_IRQ_GROUP_REBOOTED, BIT(priv->rproc_pt), false);
+ imx_scu_irq_unregister_notifier(&priv->rproc_nb);
+}
+
+static int imx_rproc_partition_notify(struct notifier_block *nb,
+ unsigned long event, void *group)
+{
+ struct imx_rproc *priv = container_of(nb, struct imx_rproc, rproc_nb);
+
+ /* Ignore other irqs */
+ if (!((event & BIT(priv->rproc_pt)) && (*(u8 *)group == IMX_SC_IRQ_GROUP_REBOOTED)))
+ return 0;
+
+ rproc_report_crash(priv->rproc, RPROC_WATCHDOG);
+
+ pr_info("Partition%d reset!\n", priv->rproc_pt);
+
+ return 0;
+}
+
static int imx_rproc_detect_mode(struct imx_rproc *priv)
{
struct regmap_config config = { .name = "imx-rproc" };
@@ -689,6 +745,7 @@ static int imx_rproc_detect_mode(struct imx_rproc *priv)
struct arm_smccc_res res;
int ret;
u32 val;
+ u8 pt;
switch (dcfg->method) {
case IMX_RPROC_NONE:
@@ -699,6 +756,51 @@ static int imx_rproc_detect_mode(struct imx_rproc *priv)
if (res.a0)
priv->rproc->state = RPROC_DETACHED;
return 0;
+ case IMX_RPROC_SCU_API:
+ ret = imx_scu_get_handle(&priv->ipc_handle);
+ if (ret)
+ return ret;
+ ret = of_property_read_u32(dev->of_node, "fsl,resource-id", &priv->rsrc_id);
+ if (ret) {
+ dev_err(dev, "No fsl,resource-id property\n");
+ return ret;
+ }
+
+ /*
+ * If Mcore resource is not owned by Acore partition, It is kicked by ROM,
+ * and Linux could only do IPC with Mcore and nothing else.
+ */
+ if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id))
+ return 0;
+
+ priv->rproc->state = RPROC_DETACHED;
+ priv->rproc->recovery_disabled = true;
+
+ /* Get partition id and enable irq in SCFW */
+ ret = imx_sc_rm_get_resource_owner(priv->ipc_handle, priv->rsrc_id, &pt);
+ if (ret) {
+ dev_err(dev, "not able to get resource owner\n");
+ return ret;
+ }
+
+ priv->rproc_pt = pt;
+ priv->rproc_nb.notifier_call = imx_rproc_partition_notify;
+
+ ret = imx_scu_irq_register_notifier(&priv->rproc_nb);
+ if (ret) {
+ dev_warn(dev, "register scu notifier failed.\n");
+ return ret;
+ }
+
+ ret = imx_scu_irq_group_enable(IMX_SC_IRQ_GROUP_REBOOTED, BIT(priv->rproc_pt),
+ true);
+ if (ret) {
+ imx_scu_irq_unregister_notifier(&priv->rproc_nb);
+ dev_warn(dev, "Enable irq failed.\n");
+ return ret;
+ }
+
+ return 0;
default:
break;
}
@@ -803,7 +905,7 @@ static int imx_rproc_probe(struct platform_device *pdev)
ret = imx_rproc_clk_enable(priv);
if (ret)
- goto err_put_mbox;
+ goto err_put_scu;
INIT_WORK(&priv->rproc_work, imx_rproc_vq_work);
@@ -820,6 +922,8 @@ static int imx_rproc_probe(struct platform_device *pdev)
err_put_clk:
clk_disable_unprepare(priv->clk);
+err_put_scu:
+ imx_rproc_put_scu(rproc);
err_put_mbox:
imx_rproc_free_mbox(rproc);
err_put_wkq:
@@ -837,6 +941,7 @@ static int imx_rproc_remove(struct platform_device *pdev)
clk_disable_unprepare(priv->clk);
rproc_del(rproc);
+ imx_rproc_put_scu(rproc);
imx_rproc_free_mbox(rproc);
destroy_workqueue(priv->workqueue);
rproc_free(rproc);
@@ -852,6 +957,7 @@ static const struct of_device_id imx_rproc_of_match[] = {
{ .compatible = "fsl,imx8mm-cm4", .data = &imx_rproc_cfg_imx8mq },
{ .compatible = "fsl,imx8mn-cm7", .data = &imx_rproc_cfg_imx8mn },
{ .compatible = "fsl,imx8mp-cm7", .data = &imx_rproc_cfg_imx8mn },
+ { .compatible = "fsl,imx8qxp-cm4", .data = &imx_rproc_cfg_imx8qxp },
{ .compatible = "fsl,imx8ulp-cm33", .data = &imx_rproc_cfg_imx8ulp },
{ .compatible = "fsl,imx93-cm33", .data = &imx_rproc_cfg_imx93 },
{},
--
2.37.1
From: Peng Fan <[email protected]>
It is possible that when remote processor crash, the communication
channel will be broken with garbage value in mailbox, such as
when Linux is issuing a message through mailbox, remote processor
crashes, we need free & rebuild the mailbox channels to make sure
no garbage value in mailbox channels.
So move the request/free to start/stop for managing remote procesosr in
Linux, move to attach/detach for remote processor is out of control of
Linux.
Previous, we just request mbox when attach for CM4 boot early before
Linux, but if mbox defer probe, remoteproc core will do resource cleanup
and corrupt resource table for later probe.
So move request mbox ealier and still keep mbox request when attach
for self recovery case, but keep a check when request/free mbox.
Signed-off-by: Peng Fan <[email protected]>
---
drivers/remoteproc/imx_rproc.c | 26 +++++++++++++++++++++++---
1 file changed, 23 insertions(+), 3 deletions(-)
diff --git a/drivers/remoteproc/imx_rproc.c b/drivers/remoteproc/imx_rproc.c
index 8175c4c9aa22..bece44b46719 100644
--- a/drivers/remoteproc/imx_rproc.c
+++ b/drivers/remoteproc/imx_rproc.c
@@ -84,6 +84,8 @@ struct imx_rproc_mem {
#define ATT_CORE_MASK 0xffff
#define ATT_CORE(I) BIT((I))
+static int imx_rproc_xtr_mbox_init(struct rproc *rproc);
+static void imx_rproc_free_mbox(struct rproc *rproc);
static int imx_rproc_detach_pd(struct rproc *rproc);
struct imx_rproc {
@@ -357,6 +359,10 @@ static int imx_rproc_start(struct rproc *rproc)
struct arm_smccc_res res;
int ret;
+ ret = imx_rproc_xtr_mbox_init(rproc);
+ if (ret)
+ return ret;
+
switch (dcfg->method) {
case IMX_RPROC_MMIO:
ret = regmap_update_bits(priv->regmap, dcfg->src_reg, dcfg->src_mask,
@@ -407,6 +413,8 @@ static int imx_rproc_stop(struct rproc *rproc)
if (ret)
dev_err(dev, "Failed to stop remote core\n");
+ else
+ imx_rproc_free_mbox(rproc);
return ret;
}
@@ -592,7 +600,7 @@ static void imx_rproc_kick(struct rproc *rproc, int vqid)
static int imx_rproc_attach(struct rproc *rproc)
{
- return 0;
+ return imx_rproc_xtr_mbox_init(rproc);
}
static struct resource_table *imx_rproc_get_loaded_rsc_table(struct rproc *rproc, size_t *table_sz)
@@ -720,6 +728,9 @@ static int imx_rproc_xtr_mbox_init(struct rproc *rproc)
struct device *dev = priv->dev;
struct mbox_client *cl;
+ if (priv->tx_ch && priv->rx_ch)
+ return 0;
+
if (!of_get_property(dev->of_node, "mbox-names", NULL))
return 0;
@@ -749,8 +760,15 @@ static void imx_rproc_free_mbox(struct rproc *rproc)
{
struct imx_rproc *priv = rproc->priv;
- mbox_free_channel(priv->tx_ch);
- mbox_free_channel(priv->rx_ch);
+ if (priv->tx_ch) {
+ mbox_free_channel(priv->tx_ch);
+ priv->tx_ch = NULL;
+ }
+
+ if (priv->rx_ch) {
+ mbox_free_channel(priv->rx_ch);
+ priv->rx_ch = NULL;
+ }
}
static void imx_rproc_put_scu(struct rproc *rproc)
@@ -779,6 +797,8 @@ static int imx_rproc_partition_notify(struct notifier_block *nb,
if (!((event & BIT(priv->rproc_pt)) && (*(u8 *)group == IMX_SC_IRQ_GROUP_REBOOTED)))
return 0;
+ imx_rproc_free_mbox(priv->rproc);
+
rproc_report_crash(priv->rproc, RPROC_WATCHDOG);
pr_info("Partition%d reset!\n", priv->rproc_pt);
--
2.37.1
From: Peng Fan <[email protected]>
Add i.MX8QM compatible
Reviewed-by: Rob Herring <[email protected]>
Signed-off-by: Peng Fan <[email protected]>
---
Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml | 1 +
1 file changed, 1 insertion(+)
diff --git a/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml b/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml
index 70322e57b6ff..64e783234e38 100644
--- a/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml
+++ b/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml
@@ -23,6 +23,7 @@ properties:
- fsl,imx8mp-cm7
- fsl,imx8mq-cm4
- fsl,imx8qxp-cm4
+ - fsl,imx8qm-cm4
- fsl,imx8ulp-cm33
- fsl,imx93-cm33
--
2.37.1
Hi Mathieu,
This patch is an incremental patch that could ignored, if you not have
time or think this patch is not in good shape. Patch 1-6 is
to achieve same as V5 patchset [1].
[1]https://lore.kernel.org/linux-remoteproc/DU0PR04MB941763C3EBF8AE9A9CD69445884C9@DU0PR04MB9417.eurprd04.prod.outlook.com/T/#md2af773bc7f251efbe8ca011b5a70ba2193b4c01
Thanks,
Peng.
On 9/29/2022 2:17 PM, Peng Fan (OSS) wrote:
> From: Peng Fan <[email protected]>
>
> i.MX8QM/QXP M4 could recover without help from Linux, so to support it:
> - enable feature RPROC_FEAT_ATTACH_ON_RECOVERY
> - add detach hook
> - Since we have detach callback, we could move the free mbox operation
> from partition reset notify to detach callback.
>
> Signed-off-by: Peng Fan <[email protected]>
> ---
> drivers/remoteproc/imx_rproc.c | 22 +++++++++++++++++++---
> 1 file changed, 19 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/remoteproc/imx_rproc.c b/drivers/remoteproc/imx_rproc.c
> index bece44b46719..4057d6f33813 100644
> --- a/drivers/remoteproc/imx_rproc.c
> +++ b/drivers/remoteproc/imx_rproc.c
> @@ -603,6 +603,22 @@ static int imx_rproc_attach(struct rproc *rproc)
> return imx_rproc_xtr_mbox_init(rproc);
> }
>
> +static int imx_rproc_detach(struct rproc *rproc)
> +{
> + struct imx_rproc *priv = rproc->priv;
> + const struct imx_rproc_dcfg *dcfg = priv->dcfg;
> +
> + if (dcfg->method != IMX_RPROC_SCU_API)
> + return -EOPNOTSUPP;
> +
> + if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id))
> + return -EOPNOTSUPP;
> +
> + imx_rproc_free_mbox(rproc);
> +
> + return 0;
> +}
> +
> static struct resource_table *imx_rproc_get_loaded_rsc_table(struct rproc *rproc, size_t *table_sz)
> {
> struct imx_rproc *priv = rproc->priv;
> @@ -618,6 +634,7 @@ static struct resource_table *imx_rproc_get_loaded_rsc_table(struct rproc *rproc
> static const struct rproc_ops imx_rproc_ops = {
> .prepare = imx_rproc_prepare,
> .attach = imx_rproc_attach,
> + .detach = imx_rproc_detach,
> .start = imx_rproc_start,
> .stop = imx_rproc_stop,
> .kick = imx_rproc_kick,
> @@ -797,8 +814,6 @@ static int imx_rproc_partition_notify(struct notifier_block *nb,
> if (!((event & BIT(priv->rproc_pt)) && (*(u8 *)group == IMX_SC_IRQ_GROUP_REBOOTED)))
> return 0;
>
> - imx_rproc_free_mbox(priv->rproc);
> -
> rproc_report_crash(priv->rproc, RPROC_WATCHDOG);
>
> pr_info("Partition%d reset!\n", priv->rproc_pt);
> @@ -916,7 +931,8 @@ static int imx_rproc_detect_mode(struct imx_rproc *priv)
> }
>
> priv->rproc->state = RPROC_DETACHED;
> - priv->rproc->recovery_disabled = true;
> + priv->rproc->recovery_disabled = false;
> + rproc_set_feature(priv->rproc, RPROC_FEAT_ATTACH_ON_RECOVERY);
>
> /* Get partition id and enable irq in SCFW */
> ret = imx_sc_rm_get_resource_owner(priv->ipc_handle, priv->rsrc_id, &pt);
On Thu, Sep 29, 2022 at 02:17:00PM +0800, Peng Fan (OSS) wrote:
> From: Peng Fan <[email protected]>
>
> When M4 is kicked by SCFW, M4 runs in its own hardware partition, Linux
> could only do IPC with M4, it could not start, stop, update image.
>
> We disable recovery reboot when M4 is managed by SCFW, because
> remoteproc core still not support M4 auto-recovery without loading
> image.
>
> Signed-off-by: Peng Fan <[email protected]>
> ---
> drivers/remoteproc/imx_rproc.c | 108 ++++++++++++++++++++++++++++++++-
> 1 file changed, 107 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/remoteproc/imx_rproc.c b/drivers/remoteproc/imx_rproc.c
> index 7cc4fd207e2d..a56aecae00c6 100644
> --- a/drivers/remoteproc/imx_rproc.c
> +++ b/drivers/remoteproc/imx_rproc.c
> @@ -6,6 +6,7 @@
> #include <linux/arm-smccc.h>
> #include <linux/clk.h>
> #include <linux/err.h>
> +#include <linux/firmware/imx/sci.h>
> #include <linux/interrupt.h>
> #include <linux/kernel.h>
> #include <linux/mailbox_client.h>
> @@ -59,6 +60,8 @@
> #define IMX_SIP_RPROC_STARTED 0x01
> #define IMX_SIP_RPROC_STOP 0x02
>
> +#define IMX_SC_IRQ_GROUP_REBOOTED 5
> +
> /**
> * struct imx_rproc_mem - slim internal memory structure
> * @cpu_addr: MPU virtual address of the memory region
> @@ -89,6 +92,10 @@ struct imx_rproc {
> struct work_struct rproc_work;
> struct workqueue_struct *workqueue;
> void __iomem *rsc_table;
> + struct imx_sc_ipc *ipc_handle;
> + struct notifier_block rproc_nb;
> + u32 rproc_pt; /* partition id */
> + u32 rsrc_id; /* resource id */
> };
>
> static const struct imx_rproc_att imx_rproc_att_imx93[] = {
> @@ -117,6 +124,18 @@ static const struct imx_rproc_att imx_rproc_att_imx93[] = {
> { 0xD0000000, 0xa0000000, 0x10000000, 0 },
> };
>
> +static const struct imx_rproc_att imx_rproc_att_imx8qxp[] = {
> + { 0x08000000, 0x08000000, 0x10000000, 0 },
> + /* TCML/U */
> + { 0x1FFE0000, 0x34FE0000, 0x00040000, ATT_OWN | ATT_IOMEM },
> + /* OCRAM(Low 96KB) */
> + { 0x21000000, 0x00100000, 0x00018000, 0 },
> + /* OCRAM */
> + { 0x21100000, 0x00100000, 0x00040000, 0 },
> + /* DDR (Data) */
> + { 0x80000000, 0x80000000, 0x60000000, 0 },
> +};
> +
> static const struct imx_rproc_att imx_rproc_att_imx8mn[] = {
> /* dev addr , sys addr , size , flags */
> /* ITCM */
> @@ -255,6 +274,12 @@ static const struct imx_rproc_dcfg imx_rproc_cfg_imx8mq = {
> .method = IMX_RPROC_MMIO,
> };
>
> +static const struct imx_rproc_dcfg imx_rproc_cfg_imx8qxp = {
> + .att = imx_rproc_att_imx8qxp,
> + .att_size = ARRAY_SIZE(imx_rproc_att_imx8qxp),
> + .method = IMX_RPROC_SCU_API,
> +};
> +
> static const struct imx_rproc_dcfg imx_rproc_cfg_imx8ulp = {
> .att = imx_rproc_att_imx8ulp,
> .att_size = ARRAY_SIZE(imx_rproc_att_imx8ulp),
> @@ -680,6 +705,37 @@ static void imx_rproc_free_mbox(struct rproc *rproc)
> mbox_free_channel(priv->rx_ch);
> }
>
> +static void imx_rproc_put_scu(struct rproc *rproc)
> +{
> + struct imx_rproc *priv = rproc->priv;
> + const struct imx_rproc_dcfg *dcfg = priv->dcfg;
> +
> + if (dcfg->method != IMX_RPROC_SCU_API)
> + return;
> +
> + if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id))
Yes, now it makes sense.
> + return;
> +
> + imx_scu_irq_group_enable(IMX_SC_IRQ_GROUP_REBOOTED, BIT(priv->rproc_pt), false);
> + imx_scu_irq_unregister_notifier(&priv->rproc_nb);
> +}
> +
> +static int imx_rproc_partition_notify(struct notifier_block *nb,
> + unsigned long event, void *group)
> +{
> + struct imx_rproc *priv = container_of(nb, struct imx_rproc, rproc_nb);
> +
> + /* Ignore other irqs */
> + if (!((event & BIT(priv->rproc_pt)) && (*(u8 *)group == IMX_SC_IRQ_GROUP_REBOOTED)))
> + return 0;
> +
> + rproc_report_crash(priv->rproc, RPROC_WATCHDOG);
> +
> + pr_info("Partition%d reset!\n", priv->rproc_pt);
> +
> + return 0;
> +}
> +
> static int imx_rproc_detect_mode(struct imx_rproc *priv)
> {
> struct regmap_config config = { .name = "imx-rproc" };
> @@ -689,6 +745,7 @@ static int imx_rproc_detect_mode(struct imx_rproc *priv)
> struct arm_smccc_res res;
> int ret;
> u32 val;
> + u8 pt;
>
> switch (dcfg->method) {
> case IMX_RPROC_NONE:
> @@ -699,6 +756,51 @@ static int imx_rproc_detect_mode(struct imx_rproc *priv)
> if (res.a0)
> priv->rproc->state = RPROC_DETACHED;
> return 0;
> + case IMX_RPROC_SCU_API:
> + ret = imx_scu_get_handle(&priv->ipc_handle);
> + if (ret)
> + return ret;
> + ret = of_property_read_u32(dev->of_node, "fsl,resource-id", &priv->rsrc_id);
> + if (ret) {
> + dev_err(dev, "No fsl,resource-id property\n");
> + return ret;
> + }
> +
> + /*
> + * If Mcore resource is not owned by Acore partition, It is kicked by ROM,
> + * and Linux could only do IPC with Mcore and nothing else.
> + */
> + if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id))
> + return 0;
> +
> + priv->rproc->state = RPROC_DETACHED;
> + priv->rproc->recovery_disabled = true;
> +
> + /* Get partition id and enable irq in SCFW */
> + ret = imx_sc_rm_get_resource_owner(priv->ipc_handle, priv->rsrc_id, &pt);
> + if (ret) {
> + dev_err(dev, "not able to get resource owner\n");
> + return ret;
> + }
> +
> + priv->rproc_pt = pt;
> + priv->rproc_nb.notifier_call = imx_rproc_partition_notify;
> +
> + ret = imx_scu_irq_register_notifier(&priv->rproc_nb);
> + if (ret) {
> + dev_warn(dev, "register scu notifier failed.\n");
Either make this a dev_err() or remove entirely.
> + return ret;
> + }
> +
> + ret = imx_scu_irq_group_enable(IMX_SC_IRQ_GROUP_REBOOTED, BIT(priv->rproc_pt),
> + true);
> + if (ret) {
> + imx_scu_irq_unregister_notifier(&priv->rproc_nb);
> + dev_warn(dev, "Enable irq failed.\n");
Same here.
With the above:
Reviewed-by: Mathieu Poirier <[email protected]>
> + return ret;
> + }
> +
> + return 0;
> default:
> break;
> }
> @@ -803,7 +905,7 @@ static int imx_rproc_probe(struct platform_device *pdev)
>
> ret = imx_rproc_clk_enable(priv);
> if (ret)
> - goto err_put_mbox;
> + goto err_put_scu;
>
> INIT_WORK(&priv->rproc_work, imx_rproc_vq_work);
>
> @@ -820,6 +922,8 @@ static int imx_rproc_probe(struct platform_device *pdev)
>
> err_put_clk:
> clk_disable_unprepare(priv->clk);
> +err_put_scu:
> + imx_rproc_put_scu(rproc);
> err_put_mbox:
> imx_rproc_free_mbox(rproc);
> err_put_wkq:
> @@ -837,6 +941,7 @@ static int imx_rproc_remove(struct platform_device *pdev)
>
> clk_disable_unprepare(priv->clk);
> rproc_del(rproc);
> + imx_rproc_put_scu(rproc);
> imx_rproc_free_mbox(rproc);
> destroy_workqueue(priv->workqueue);
> rproc_free(rproc);
> @@ -852,6 +957,7 @@ static const struct of_device_id imx_rproc_of_match[] = {
> { .compatible = "fsl,imx8mm-cm4", .data = &imx_rproc_cfg_imx8mq },
> { .compatible = "fsl,imx8mn-cm7", .data = &imx_rproc_cfg_imx8mn },
> { .compatible = "fsl,imx8mp-cm7", .data = &imx_rproc_cfg_imx8mn },
> + { .compatible = "fsl,imx8qxp-cm4", .data = &imx_rproc_cfg_imx8qxp },
> { .compatible = "fsl,imx8ulp-cm33", .data = &imx_rproc_cfg_imx8ulp },
> { .compatible = "fsl,imx93-cm33", .data = &imx_rproc_cfg_imx93 },
> {},
> --
> 2.37.1
>
On Thu, Sep 29, 2022 at 02:17:01PM +0800, Peng Fan (OSS) wrote:
> From: Peng Fan <[email protected]>
>
> When M4 is in the same hardware partition with Cortex-A, it
> could be start/stop by Linux.
>
> Added power domain to make sure M4 could run, it requires several power
> domains to work. Make clk always optional for i.MX8QXP, because
> SCFW handles it when power up M4 core.
>
> Signed-off-by: Peng Fan <[email protected]>
> ---
> drivers/remoteproc/imx_rproc.c | 88 ++++++++++++++++++++++++++++++++--
> 1 file changed, 85 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/remoteproc/imx_rproc.c b/drivers/remoteproc/imx_rproc.c
> index a56aecae00c6..7d2e11aed0aa 100644
> --- a/drivers/remoteproc/imx_rproc.c
> +++ b/drivers/remoteproc/imx_rproc.c
> @@ -16,6 +16,7 @@
> #include <linux/of_reserved_mem.h>
> #include <linux/of_device.h>
> #include <linux/platform_device.h>
> +#include <linux/pm_domain.h>
> #include <linux/regmap.h>
> #include <linux/remoteproc.h>
> #include <linux/workqueue.h>
> @@ -79,6 +80,8 @@ struct imx_rproc_mem {
> #define ATT_OWN BIT(1)
> #define ATT_IOMEM BIT(2)
>
> +static int imx_rproc_detach_pd(struct rproc *rproc);
> +
> struct imx_rproc {
> struct device *dev;
> struct regmap *regmap;
> @@ -96,6 +99,10 @@ struct imx_rproc {
> struct notifier_block rproc_nb;
> u32 rproc_pt; /* partition id */
> u32 rsrc_id; /* resource id */
> + u32 entry; /* cpu start address */
> + int num_pd;
> + struct device **pd_dev;
> + struct device_link **pd_dev_link;
> };
>
> static const struct imx_rproc_att imx_rproc_att_imx93[] = {
> @@ -335,6 +342,9 @@ static int imx_rproc_start(struct rproc *rproc)
> arm_smccc_smc(IMX_SIP_RPROC, IMX_SIP_RPROC_START, 0, 0, 0, 0, 0, 0, &res);
> ret = res.a0;
> break;
> + case IMX_RPROC_SCU_API:
> + ret = imx_sc_pm_cpu_start(priv->ipc_handle, priv->rsrc_id, true, priv->entry);
> + break;
> default:
> return -EOPNOTSUPP;
> }
> @@ -364,6 +374,9 @@ static int imx_rproc_stop(struct rproc *rproc)
> if (res.a1)
> dev_info(dev, "Not in wfi, force stopped\n");
> break;
> + case IMX_RPROC_SCU_API:
> + ret = imx_sc_pm_cpu_start(priv->ipc_handle, priv->rsrc_id, false, priv->entry);
> + break;
> default:
> return -EOPNOTSUPP;
> }
> @@ -713,8 +726,10 @@ static void imx_rproc_put_scu(struct rproc *rproc)
> if (dcfg->method != IMX_RPROC_SCU_API)
> return;
>
> - if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id))
> + if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id)) {
> + imx_rproc_detach_pd(rproc);
> return;
> + }
>
> imx_scu_irq_group_enable(IMX_SC_IRQ_GROUP_REBOOTED, BIT(priv->rproc_pt), false);
> imx_scu_irq_unregister_notifier(&priv->rproc_nb);
> @@ -736,6 +751,69 @@ static int imx_rproc_partition_notify(struct notifier_block *nb,
> return 0;
> }
>
> +static int imx_rproc_attach_pd(struct imx_rproc *priv)
> +{
> + struct device *dev = priv->dev;
> + int ret, i;
> +
> + priv->num_pd = of_count_phandle_with_args(dev->of_node, "power-domains",
> + "#power-domain-cells");
> + if (priv->num_pd <= 1)
What happens if we get 1 power domains? How is it handled? There should be a
comment explaining what is happening here.
> + return 0;
> +
> + priv->pd_dev = devm_kmalloc_array(dev, priv->num_pd, sizeof(*priv->pd_dev), GFP_KERNEL);
> + if (!priv->pd_dev)
> + return -ENOMEM;
> +
> + priv->pd_dev_link = devm_kmalloc_array(dev, priv->num_pd, sizeof(*priv->pd_dev_link),
> + GFP_KERNEL);
> +
> + if (!priv->pd_dev_link)
> + return -ENOMEM;
> +
> + for (i = 0; i < priv->num_pd; i++) {
> + priv->pd_dev[i] = dev_pm_domain_attach_by_id(dev, i);
> + if (IS_ERR(priv->pd_dev[i])) {
> + ret = PTR_ERR(priv->pd_dev[i]);
> + goto detach_pd;
> + }
> +
> + priv->pd_dev_link[i] = device_link_add(dev, priv->pd_dev[i], DL_FLAG_STATELESS |
> + DL_FLAG_PM_RUNTIME | DL_FLAG_RPM_ACTIVE);
> + if (!priv->pd_dev_link[i]) {
> + dev_pm_domain_detach(priv->pd_dev[i], false);
> + ret = -EINVAL;
> + goto detach_pd;
> + }
> + }
> +
> + return 0;
> +
> +detach_pd:
> + while (--i >= 0) {
> + device_link_del(priv->pd_dev_link[i]);
> + dev_pm_domain_detach(priv->pd_dev[i], false);
> + }
> +
> + return ret;
> +}
> +
> +static int imx_rproc_detach_pd(struct rproc *rproc)
> +{
> + struct imx_rproc *priv = rproc->priv;
> + int i;
> +
> + if (priv->num_pd <= 1)
Same here.
With the above:
Reviewed-by: Mathieu Poirier <[email protected]>
> + return 0;
> +
> + for (i = 0; i < priv->num_pd; i++) {
> + device_link_del(priv->pd_dev_link[i]);
> + dev_pm_domain_detach(priv->pd_dev[i], false);
> + }
> +
> + return 0;
> +}
> +
> static int imx_rproc_detect_mode(struct imx_rproc *priv)
> {
> struct regmap_config config = { .name = "imx-rproc" };
> @@ -770,8 +848,12 @@ static int imx_rproc_detect_mode(struct imx_rproc *priv)
> * If Mcore resource is not owned by Acore partition, It is kicked by ROM,
> * and Linux could only do IPC with Mcore and nothing else.
> */
> - if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id))
> - return 0;
> + if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id)) {
> + if (of_property_read_u32(dev->of_node, "fsl,entry-address", &priv->entry))
> + return -EINVAL;
> +
> + return imx_rproc_attach_pd(priv);
> + }
>
> priv->rproc->state = RPROC_DETACHED;
> priv->rproc->recovery_disabled = true;
> --
> 2.37.1
>
On Thu, Sep 29, 2022 at 02:17:03PM +0800, Peng Fan (OSS) wrote:
> From: Peng Fan <[email protected]>
>
> It is possible that when remote processor crash, the communication
> channel will be broken with garbage value in mailbox, such as
> when Linux is issuing a message through mailbox, remote processor
> crashes, we need free & rebuild the mailbox channels to make sure
> no garbage value in mailbox channels.
>
> So move the request/free to start/stop for managing remote procesosr in
> Linux, move to attach/detach for remote processor is out of control of
> Linux.
>
> Previous, we just request mbox when attach for CM4 boot early before
> Linux, but if mbox defer probe, remoteproc core will do resource cleanup
> and corrupt resource table for later probe.
>
> So move request mbox ealier and still keep mbox request when attach
> for self recovery case, but keep a check when request/free mbox.
>
> Signed-off-by: Peng Fan <[email protected]>
> ---
> drivers/remoteproc/imx_rproc.c | 26 +++++++++++++++++++++++---
> 1 file changed, 23 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/remoteproc/imx_rproc.c b/drivers/remoteproc/imx_rproc.c
> index 8175c4c9aa22..bece44b46719 100644
> --- a/drivers/remoteproc/imx_rproc.c
> +++ b/drivers/remoteproc/imx_rproc.c
> @@ -84,6 +84,8 @@ struct imx_rproc_mem {
> #define ATT_CORE_MASK 0xffff
> #define ATT_CORE(I) BIT((I))
>
> +static int imx_rproc_xtr_mbox_init(struct rproc *rproc);
> +static void imx_rproc_free_mbox(struct rproc *rproc);
> static int imx_rproc_detach_pd(struct rproc *rproc);
>
> struct imx_rproc {
> @@ -357,6 +359,10 @@ static int imx_rproc_start(struct rproc *rproc)
> struct arm_smccc_res res;
> int ret;
>
> + ret = imx_rproc_xtr_mbox_init(rproc);
> + if (ret)
> + return ret;
> +
> switch (dcfg->method) {
> case IMX_RPROC_MMIO:
> ret = regmap_update_bits(priv->regmap, dcfg->src_reg, dcfg->src_mask,
> @@ -407,6 +413,8 @@ static int imx_rproc_stop(struct rproc *rproc)
>
> if (ret)
> dev_err(dev, "Failed to stop remote core\n");
> + else
> + imx_rproc_free_mbox(rproc);
>
> return ret;
> }
> @@ -592,7 +600,7 @@ static void imx_rproc_kick(struct rproc *rproc, int vqid)
>
> static int imx_rproc_attach(struct rproc *rproc)
> {
> - return 0;
> + return imx_rproc_xtr_mbox_init(rproc);
> }
>
> static struct resource_table *imx_rproc_get_loaded_rsc_table(struct rproc *rproc, size_t *table_sz)
> @@ -720,6 +728,9 @@ static int imx_rproc_xtr_mbox_init(struct rproc *rproc)
> struct device *dev = priv->dev;
> struct mbox_client *cl;
>
> + if (priv->tx_ch && priv->rx_ch)
> + return 0;
> +
Why is this needed? The remoteproc core's state machine driven by rproc->state
guarantees that imx_rproc_start(), imx_rproc_stop() and imx_rproc_attach are not
called more than once.
And why is imx_rproc_xtr_mbox_init() in imx_rproc_probe() not removed?
> if (!of_get_property(dev->of_node, "mbox-names", NULL))
> return 0;
>
> @@ -749,8 +760,15 @@ static void imx_rproc_free_mbox(struct rproc *rproc)
> {
> struct imx_rproc *priv = rproc->priv;
>
> - mbox_free_channel(priv->tx_ch);
> - mbox_free_channel(priv->rx_ch);
> + if (priv->tx_ch) {
> + mbox_free_channel(priv->tx_ch);
> + priv->tx_ch = NULL;
> + }
> +
> + if (priv->rx_ch) {
> + mbox_free_channel(priv->rx_ch);
> + priv->rx_ch = NULL;
> + }
> }
>
> static void imx_rproc_put_scu(struct rproc *rproc)
> @@ -779,6 +797,8 @@ static int imx_rproc_partition_notify(struct notifier_block *nb,
> if (!((event & BIT(priv->rproc_pt)) && (*(u8 *)group == IMX_SC_IRQ_GROUP_REBOOTED)))
> return 0;
>
> + imx_rproc_free_mbox(priv->rproc);
Why putting this here when it will be called again in imx_rproc_stop()? If you
need it for the attached scenario, create imx_rproc_detach() and add it there.
> +
> rproc_report_crash(priv->rproc, RPROC_WATCHDOG);
>
> pr_info("Partition%d reset!\n", priv->rproc_pt);
> --
> 2.37.1
>
On Thu, Sep 29, 2022 at 02:17:04PM +0800, Peng Fan (OSS) wrote:
> From: Peng Fan <[email protected]>
>
> i.MX8QM/QXP M4 could recover without help from Linux, so to support it:
> - enable feature RPROC_FEAT_ATTACH_ON_RECOVERY
> - add detach hook
> - Since we have detach callback, we could move the free mbox operation
> from partition reset notify to detach callback.
>
> Signed-off-by: Peng Fan <[email protected]>
> ---
> drivers/remoteproc/imx_rproc.c | 22 +++++++++++++++++++---
> 1 file changed, 19 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/remoteproc/imx_rproc.c b/drivers/remoteproc/imx_rproc.c
> index bece44b46719..4057d6f33813 100644
> --- a/drivers/remoteproc/imx_rproc.c
> +++ b/drivers/remoteproc/imx_rproc.c
> @@ -603,6 +603,22 @@ static int imx_rproc_attach(struct rproc *rproc)
> return imx_rproc_xtr_mbox_init(rproc);
> }
>
> +static int imx_rproc_detach(struct rproc *rproc)
> +{
> + struct imx_rproc *priv = rproc->priv;
> + const struct imx_rproc_dcfg *dcfg = priv->dcfg;
> +
> + if (dcfg->method != IMX_RPROC_SCU_API)
> + return -EOPNOTSUPP;
> +
> + if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id))
> + return -EOPNOTSUPP;
> +
> + imx_rproc_free_mbox(rproc);
Which is exactly what you did here.
I really don't understand what you did in patch 6/7... Anyways, I suggest to
add support for the detached scenario and then fix the mailbox corruption
problem.
I am done reviewing this set.
> +
> + return 0;
> +}
> +
> static struct resource_table *imx_rproc_get_loaded_rsc_table(struct rproc *rproc, size_t *table_sz)
> {
> struct imx_rproc *priv = rproc->priv;
> @@ -618,6 +634,7 @@ static struct resource_table *imx_rproc_get_loaded_rsc_table(struct rproc *rproc
> static const struct rproc_ops imx_rproc_ops = {
> .prepare = imx_rproc_prepare,
> .attach = imx_rproc_attach,
> + .detach = imx_rproc_detach,
> .start = imx_rproc_start,
> .stop = imx_rproc_stop,
> .kick = imx_rproc_kick,
> @@ -797,8 +814,6 @@ static int imx_rproc_partition_notify(struct notifier_block *nb,
> if (!((event & BIT(priv->rproc_pt)) && (*(u8 *)group == IMX_SC_IRQ_GROUP_REBOOTED)))
> return 0;
>
> - imx_rproc_free_mbox(priv->rproc);
> -
> rproc_report_crash(priv->rproc, RPROC_WATCHDOG);
>
> pr_info("Partition%d reset!\n", priv->rproc_pt);
> @@ -916,7 +931,8 @@ static int imx_rproc_detect_mode(struct imx_rproc *priv)
> }
>
> priv->rproc->state = RPROC_DETACHED;
> - priv->rproc->recovery_disabled = true;
> + priv->rproc->recovery_disabled = false;
> + rproc_set_feature(priv->rproc, RPROC_FEAT_ATTACH_ON_RECOVERY);
>
> /* Get partition id and enable irq in SCFW */
> ret = imx_sc_rm_get_resource_owner(priv->ipc_handle, priv->rsrc_id, &pt);
> --
> 2.37.1
>
Hi Mathieu,
> Subject: Re: [PATCH V6 6/7] remoteproc: imx_rproc: request mbox channel
> later
>
> On Thu, Sep 29, 2022 at 02:17:03PM +0800, Peng Fan (OSS) wrote:
> > From: Peng Fan <[email protected]>
> >
> > It is possible that when remote processor crash, the communication
> > channel will be broken with garbage value in mailbox, such as when
> > Linux is issuing a message through mailbox, remote processor crashes,
> > we need free & rebuild the mailbox channels to make sure no garbage
> > value in mailbox channels.
> >
> > So move the request/free to start/stop for managing remote procesosr
> > in Linux, move to attach/detach for remote processor is out of control
> > of Linux.
> >
> > Previous, we just request mbox when attach for CM4 boot early before
> > Linux, but if mbox defer probe, remoteproc core will do resource
> > cleanup and corrupt resource table for later probe.
> >
> > So move request mbox ealier and still keep mbox request when attach
> > for self recovery case, but keep a check when request/free mbox.
> >
> > Signed-off-by: Peng Fan <[email protected]>
> > ---
> > drivers/remoteproc/imx_rproc.c | 26 +++++++++++++++++++++++---
> > 1 file changed, 23 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/remoteproc/imx_rproc.c
> > b/drivers/remoteproc/imx_rproc.c index 8175c4c9aa22..bece44b46719
> > 100644
> > --- a/drivers/remoteproc/imx_rproc.c
> > +++ b/drivers/remoteproc/imx_rproc.c
> > @@ -84,6 +84,8 @@ struct imx_rproc_mem {
> > #define ATT_CORE_MASK 0xffff
> > #define ATT_CORE(I) BIT((I))
> >
> > +static int imx_rproc_xtr_mbox_init(struct rproc *rproc); static void
> > +imx_rproc_free_mbox(struct rproc *rproc);
> > static int imx_rproc_detach_pd(struct rproc *rproc);
> >
> > struct imx_rproc {
> > @@ -357,6 +359,10 @@ static int imx_rproc_start(struct rproc *rproc)
> > struct arm_smccc_res res;
> > int ret;
> >
> > + ret = imx_rproc_xtr_mbox_init(rproc);
> > + if (ret)
> > + return ret;
> > +
> > switch (dcfg->method) {
> > case IMX_RPROC_MMIO:
> > ret = regmap_update_bits(priv->regmap, dcfg->src_reg,
> > dcfg->src_mask, @@ -407,6 +413,8 @@ static int imx_rproc_stop(struct
> > rproc *rproc)
> >
> > if (ret)
> > dev_err(dev, "Failed to stop remote core\n");
> > + else
> > + imx_rproc_free_mbox(rproc);
> >
> > return ret;
> > }
> > @@ -592,7 +600,7 @@ static void imx_rproc_kick(struct rproc *rproc,
> > int vqid)
> >
> > static int imx_rproc_attach(struct rproc *rproc) {
> > - return 0;
> > + return imx_rproc_xtr_mbox_init(rproc);
> > }
> >
> > static struct resource_table *imx_rproc_get_loaded_rsc_table(struct
> > rproc *rproc, size_t *table_sz) @@ -720,6 +728,9 @@ static int
> imx_rproc_xtr_mbox_init(struct rproc *rproc)
> > struct device *dev = priv->dev;
> > struct mbox_client *cl;
> >
> > + if (priv->tx_ch && priv->rx_ch)
> > + return 0;
> > +
>
> Why is this needed? The remoteproc core's state machine driven by rproc-
> >state guarantees that imx_rproc_start(), imx_rproc_stop() and
> imx_rproc_attach are not called more than once.
>
> And why is imx_rproc_xtr_mbox_init() in imx_rproc_probe() not removed?
I not remove it, code piece in probe function:
ret = imx_rproc_xtr_mbox_init(rproc);
if (ret)
goto err_put_wkq;
ret = imx_rproc_addr_init(priv, pdev);
if (ret) {
dev_err(dev, "failed on imx_rproc_addr_init\n");
goto err_put_mbox;
}
ret = imx_rproc_detect_mode(priv);
if (ret)
goto err_put_mbox;
>
> > if (!of_get_property(dev->of_node, "mbox-names", NULL))
> > return 0;
> >
> > @@ -749,8 +760,15 @@ static void imx_rproc_free_mbox(struct rproc
> > *rproc) {
> > struct imx_rproc *priv = rproc->priv;
> >
> > - mbox_free_channel(priv->tx_ch);
> > - mbox_free_channel(priv->rx_ch);
> > + if (priv->tx_ch) {
> > + mbox_free_channel(priv->tx_ch);
> > + priv->tx_ch = NULL;
> > + }
> > +
> > + if (priv->rx_ch) {
> > + mbox_free_channel(priv->rx_ch);
> > + priv->rx_ch = NULL;
> > + }
> > }
> >
> > static void imx_rproc_put_scu(struct rproc *rproc) @@ -779,6 +797,8
> > @@ static int imx_rproc_partition_notify(struct notifier_block *nb,
> > if (!((event & BIT(priv->rproc_pt)) && (*(u8 *)group ==
> IMX_SC_IRQ_GROUP_REBOOTED)))
> > return 0;
> >
> > + imx_rproc_free_mbox(priv->rproc);
>
> Why putting this here when it will be called again in imx_rproc_stop()? If
> you need it for the attached scenario, create imx_rproc_detach() and add it
> there.
This is when got notification from CM4, Linux free the mbox channel.
The noficiation only work when CM4 runs in separated partition, linux not
able to start/stop CM4.
Patch 7/7 attach recovery will move this out. Here is just free mbox after CM4
crash, and linux not has permission to stop CM4.
Thanks,
Peng.
>
> > +
> > rproc_report_crash(priv->rproc, RPROC_WATCHDOG);
> >
> > pr_info("Partition%d reset!\n", priv->rproc_pt);
> > --
> > 2.37.1
> >
Hi Mathieu,
> Subject: Re: [PATCH V6 7/7] remoteproc: imx_rproc: Enable attach recovery
> for i.MX8QM/QXP
>
> On Thu, Sep 29, 2022 at 02:17:04PM +0800, Peng Fan (OSS) wrote:
> > From: Peng Fan <[email protected]>
> >
> > i.MX8QM/QXP M4 could recover without help from Linux, so to support it:
> > - enable feature RPROC_FEAT_ATTACH_ON_RECOVERY
> > - add detach hook
> > - Since we have detach callback, we could move the free mbox
> > operation from partition reset notify to detach callback.
> >
> > Signed-off-by: Peng Fan <[email protected]>
> > ---
> > drivers/remoteproc/imx_rproc.c | 22 +++++++++++++++++++---
> > 1 file changed, 19 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/remoteproc/imx_rproc.c
> > b/drivers/remoteproc/imx_rproc.c index bece44b46719..4057d6f33813
> > 100644
> > --- a/drivers/remoteproc/imx_rproc.c
> > +++ b/drivers/remoteproc/imx_rproc.c
> > @@ -603,6 +603,22 @@ static int imx_rproc_attach(struct rproc *rproc)
> > return imx_rproc_xtr_mbox_init(rproc); }
> >
> > +static int imx_rproc_detach(struct rproc *rproc) {
> > + struct imx_rproc *priv = rproc->priv;
> > + const struct imx_rproc_dcfg *dcfg = priv->dcfg;
> > +
> > + if (dcfg->method != IMX_RPROC_SCU_API)
> > + return -EOPNOTSUPP;
> > +
> > + if (imx_sc_rm_is_resource_owned(priv->ipc_handle, priv->rsrc_id))
> > + return -EOPNOTSUPP;
> > +
> > + imx_rproc_free_mbox(rproc);
>
> Which is exactly what you did here.
>
> I really don't understand what you did in patch 6/7... Anyways, I suggest to
> add support for the detached scenario and then fix the mailbox corruption
> problem.
When CM4 doing attach recovery, the mbox channel needs to be released,
then attach function will request the mbox channel again.
So the flow is
CM4(attach->request mbox)->CM4 crash notify Linux[1]->Linux got notification
->Linux detach(release mbox)[2]->Linux re-attach(request mbox)
Since patch 7/7 is to support attach recovery, so I move mbox free
from [1] to [2].
>
> I am done reviewing this set.
Thanks for your time.
Thanks,
Peng.
>
> > +
> > + return 0;
> > +}
> > +
> > static struct resource_table *imx_rproc_get_loaded_rsc_table(struct
> > rproc *rproc, size_t *table_sz) {
> > struct imx_rproc *priv = rproc->priv; @@ -618,6 +634,7 @@ static
> > struct resource_table *imx_rproc_get_loaded_rsc_table(struct rproc
> > *rproc static const struct rproc_ops imx_rproc_ops = {
> > .prepare = imx_rproc_prepare,
> > .attach = imx_rproc_attach,
> > + .detach = imx_rproc_detach,
> > .start = imx_rproc_start,
> > .stop = imx_rproc_stop,
> > .kick = imx_rproc_kick,
> > @@ -797,8 +814,6 @@ static int imx_rproc_partition_notify(struct
> notifier_block *nb,
> > if (!((event & BIT(priv->rproc_pt)) && (*(u8 *)group ==
> IMX_SC_IRQ_GROUP_REBOOTED)))
> > return 0;
> >
> > - imx_rproc_free_mbox(priv->rproc);
> > -
> > rproc_report_crash(priv->rproc, RPROC_WATCHDOG);
> >
> > pr_info("Partition%d reset!\n", priv->rproc_pt); @@ -916,7 +931,8
> @@
> > static int imx_rproc_detect_mode(struct imx_rproc *priv)
> > }
> >
> > priv->rproc->state = RPROC_DETACHED;
> > - priv->rproc->recovery_disabled = true;
> > + priv->rproc->recovery_disabled = false;
> > + rproc_set_feature(priv->rproc,
> RPROC_FEAT_ATTACH_ON_RECOVERY);
> >
> > /* Get partition id and enable irq in SCFW */
> > ret = imx_sc_rm_get_resource_owner(priv->ipc_handle,
> priv->rsrc_id,
> > &pt);
> > --
> > 2.37.1
> >