[PATCH 00/29] MHI changes for v5.11

[PATCH 00/29] MHI changes for v5.11

[Manivannan Sadhasivam]

Hi Greg,

Here is the MHI patch series for v5.11. It was a busy cycle for us and so far we
have accumulated almost 29 patches. Most of the patches are cleanups and fixes
but there are some noticeable changes too:

1. Loic finally removed the auto-start option from the channel parameters of the
MHI controller. It is the duty of the client drivers like qrtr to start/stop the
channels when required, so we decided to remove this option. As a side effect,
we changed the qrtr driver to start the channels during its probe and removed
the auto-start option from ath11k controller.

**NOTE** Since these changes spawns both MHI and networking trees, the patches
are maintained in an immutable branch [1] and pulled into both mhi-next and
ath11k-next branches. The networking patches got acks from ath11k and networking
maintainers as well.

2. Loic added a generic MHI pci controller driver. This driver will be used by
the PCI based Qualcomm modems like SDX55 and exposes channels such as QMI,
IP_HW0, IPCR etc...

3. Loic fixed the MHI device hierarchy by maintaining the correct parent child
relationships. Earlier all MHI devices lived in the same level under the parent
device like PCIe. But now, the MHI devices belonging to channels will become the
children of controller MHI device.

4. Finally Loic also improved the MHI device naming by using indexed names such
as mhi0, mhi1, etc... This will break the userspace applications depending on
the old naming convention but since the only one user so far is Jeff Hugo's AI
accelerator apps, we decided to make this change now itself with his agreement.

5. Bhaumik fixed the qrtr driver by stopping the channels during remove. This
patch also got ack from networking maintainer and we decided to take it through
MHI tree (via immutable branch) since we already had a qrtr change.

Please consider applying!

Thanks,
Mani

[1] https://git.kernel.org/pub/scm/linux/kernel/git/mani/mhi.git/log/?h=mhi-ath11k-immutable

Bhaumik Bhatt (19):
  bus: mhi: core: Remove unnecessary counter from mhi_firmware_copy()
  bus: mhi: core: Add missing EXPORT_SYMBOL for mhi_get_mhi_state()
  bus: mhi: core: Expose mhi_get_exec_env() API for controllers
  bus: mhi: core: Remove unused mhi_fw_load_worker() declaration
  bus: mhi: core: Rename RDDM download function to use proper words
  bus: mhi: core: Skip RDDM download for unknown execution environment
  bus: mhi: core: Use appropriate names for firmware load functions
  bus: mhi: core: Move to using high priority workqueue
  bus: mhi: core: Skip device wake in error or shutdown states
  bus: mhi: core: Move to SYS_ERROR regardless of RDDM capability
  bus: mhi: core: Prevent sending multiple RDDM entry callbacks
  bus: mhi: core: Move to an error state on any firmware load failure
  bus: mhi: core: Use appropriate label in firmware load handler API
  bus: mhi: core: Move to an error state on mission mode failure
  bus: mhi: core: Check for IRQ availability during registration
  bus: mhi: core: Separate system error and power down handling
  bus: mhi: core: Mark and maintain device states early on after power
    down
  bus: mhi: core: Remove MHI event ring IRQ handlers when powering down
  net: qrtr: Unprepare MHI channels during remove

Carl Yin (1):
  bus: mhi: core: Fix null pointer access when parsing MHI configuration

Dan Carpenter (1):
  bus: mhi: core: Fix error handling in mhi_register_controller()

Jeffrey Hugo (1):
  bus: mhi: core: fix potential operator-precedence with BHI macros

Loic Poulain (7):
  bus: mhi: Remove auto-start option
  net: qrtr: Start MHI channels during init
  bus: mhi: Add MHI PCI support for WWAN modems
  bus: mhi: Fix channel close issue on driver remove
  bus: mhi: core: Indexed MHI controller name
  bus: mhi: core: Fix device hierarchy
  mhi: pci_generic: Fix implicit conversion warning

-- 
2.25.1

[PATCH 01/29] bus: mhi: Remove auto-start option

[Manivannan Sadhasivam]

From: Loic Poulain <loic.poulain@linaro.org>

There is really no point having an auto-start for channels.
This is confusing for the device drivers, some have to enable the
channels, others don't have... and waste resources (e.g. pre allocated
buffers) that may never be used.

This is really up to the MHI device(channel) driver to manage the state
of its channels.

While at it, let's also remove the auto-start option from ath11k mhi
controller.

Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Acked-by: Kalle Valo <kvalo@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
[mani: clubbed ath11k change]
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/init.c           | 9 ---------
 drivers/bus/mhi/core/internal.h       | 1 -
 drivers/net/wireless/ath/ath11k/mhi.c | 4 ----
 include/linux/mhi.h                   | 2 --
 4 files changed, 16 deletions(-)

diff --git a/drivers/bus/mhi/core/init.c b/drivers/bus/mhi/core/init.c
index 0ffdebde8265..381fdea2eb9f 100644
--- a/drivers/bus/mhi/core/init.c
+++ b/drivers/bus/mhi/core/init.c
@@ -758,7 +758,6 @@ static int parse_ch_cfg(struct mhi_controller *mhi_cntrl,
 		mhi_chan->offload_ch = ch_cfg->offload_channel;
 		mhi_chan->db_cfg.reset_req = ch_cfg->doorbell_mode_switch;
 		mhi_chan->pre_alloc = ch_cfg->auto_queue;
-		mhi_chan->auto_start = ch_cfg->auto_start;
 
 		/*
 		 * If MHI host allocates buffers, then the channel direction
@@ -1160,11 +1159,6 @@ static int mhi_driver_probe(struct device *dev)
 			goto exit_probe;
 
 		ul_chan->xfer_cb = mhi_drv->ul_xfer_cb;
-		if (ul_chan->auto_start) {
-			ret = mhi_prepare_channel(mhi_cntrl, ul_chan);
-			if (ret)
-				goto exit_probe;
-		}
 	}
 
 	ret = -EINVAL;
@@ -1198,9 +1192,6 @@ static int mhi_driver_probe(struct device *dev)
 	if (ret)
 		goto exit_probe;
 
-	if (dl_chan && dl_chan->auto_start)
-		mhi_prepare_channel(mhi_cntrl, dl_chan);
-
 	mhi_device_put(mhi_dev);
 
 	return ret;
diff --git a/drivers/bus/mhi/core/internal.h b/drivers/bus/mhi/core/internal.h
index 7989269ddd96..33c23203c531 100644
--- a/drivers/bus/mhi/core/internal.h
+++ b/drivers/bus/mhi/core/internal.h
@@ -563,7 +563,6 @@ struct mhi_chan {
 	bool configured;
 	bool offload_ch;
 	bool pre_alloc;
-	bool auto_start;
 	bool wake_capable;
 };
 
diff --git a/drivers/net/wireless/ath/ath11k/mhi.c b/drivers/net/wireless/ath/ath11k/mhi.c
index aded9a719d51..47a1ce1bee4f 100644
--- a/drivers/net/wireless/ath/ath11k/mhi.c
+++ b/drivers/net/wireless/ath/ath11k/mhi.c
@@ -24,7 +24,6 @@ static struct mhi_channel_config ath11k_mhi_channels[] = {
 		.offload_channel = false,
 		.doorbell_mode_switch = false,
 		.auto_queue = false,
-		.auto_start = false,
 	},
 	{
 		.num = 1,
@@ -39,7 +38,6 @@ static struct mhi_channel_config ath11k_mhi_channels[] = {
 		.offload_channel = false,
 		.doorbell_mode_switch = false,
 		.auto_queue = false,
-		.auto_start = false,
 	},
 	{
 		.num = 20,
@@ -54,7 +52,6 @@ static struct mhi_channel_config ath11k_mhi_channels[] = {
 		.offload_channel = false,
 		.doorbell_mode_switch = false,
 		.auto_queue = false,
-		.auto_start = true,
 	},
 	{
 		.num = 21,
@@ -69,7 +66,6 @@ static struct mhi_channel_config ath11k_mhi_channels[] = {
 		.offload_channel = false,
 		.doorbell_mode_switch = false,
 		.auto_queue = true,
-		.auto_start = true,
 	},
 };
 
diff --git a/include/linux/mhi.h b/include/linux/mhi.h
index d4841e5a5f45..6522a4adc794 100644
--- a/include/linux/mhi.h
+++ b/include/linux/mhi.h
@@ -214,7 +214,6 @@ enum mhi_db_brst_mode {
  * @offload_channel: The client manages the channel completely
  * @doorbell_mode_switch: Channel switches to doorbell mode on M0 transition
  * @auto_queue: Framework will automatically queue buffers for DL traffic
- * @auto_start: Automatically start (open) this channel
  * @wake-capable: Channel capable of waking up the system
  */
 struct mhi_channel_config {
@@ -232,7 +231,6 @@ struct mhi_channel_config {
 	bool offload_channel;
 	bool doorbell_mode_switch;
 	bool auto_queue;
-	bool auto_start;
 	bool wake_capable;
 };
 
-- 
2.25.1

[PATCH 02/29] net: qrtr: Start MHI channels during init

[Manivannan Sadhasivam]

From: Loic Poulain <loic.poulain@linaro.org>

Start MHI device channels so that transfers can be performed.
The MHI stack does not auto-start channels anymore.

Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 net/qrtr/mhi.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/net/qrtr/mhi.c b/net/qrtr/mhi.c
index ff0c41467fc1..7100f0bac4c6 100644
--- a/net/qrtr/mhi.c
+++ b/net/qrtr/mhi.c
@@ -76,6 +76,11 @@ static int qcom_mhi_qrtr_probe(struct mhi_device *mhi_dev,
 	struct qrtr_mhi_dev *qdev;
 	int rc;
 
+	/* start channels */
+	rc = mhi_prepare_for_transfer(mhi_dev);
+	if (rc)
+		return rc;
+
 	qdev = devm_kzalloc(&mhi_dev->dev, sizeof(*qdev), GFP_KERNEL);
 	if (!qdev)
 		return -ENOMEM;
-- 
2.25.1

[PATCH 03/29] bus: mhi: core: fix potential operator-precedence with BHI macros

[Manivannan Sadhasivam]

From: Jeffrey Hugo <jhugo@codeaurora.org>

The BHI_MSMHWID and BHI_OEMPKHASH macros take a value 'n' which is
a BHI register index. If 'n' is an expression rather than a simple
value, there can be an operator precedence issue which can result
in the incorrect calculation of the register offset. Adding
parentheses around the macro parameter can prevent such issues.

Signed-off-by: Jeffrey Hugo <jhugo@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Reviewed-by: Hemant Kumar <hemantk@codeaurora.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/internal.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/bus/mhi/core/internal.h b/drivers/bus/mhi/core/internal.h
index 7989269ddd96..78e4e84d6743 100644
--- a/drivers/bus/mhi/core/internal.h
+++ b/drivers/bus/mhi/core/internal.h
@@ -153,8 +153,8 @@ extern struct bus_type mhi_bus_type;
 #define BHI_SERIALNU (0x40)
 #define BHI_SBLANTIROLLVER (0x44)
 #define BHI_NUMSEG (0x48)
-#define BHI_MSMHWID(n) (0x4C + (0x4 * n))
-#define BHI_OEMPKHASH(n) (0x64 + (0x4 * n))
+#define BHI_MSMHWID(n) (0x4C + (0x4 * (n)))
+#define BHI_OEMPKHASH(n) (0x64 + (0x4 * (n)))
 #define BHI_RSVD5 (0xC4)
 #define BHI_STATUS_MASK (0xC0000000)
 #define BHI_STATUS_SHIFT (30)
-- 
2.25.1

[PATCH 04/29] bus: mhi: Add MHI PCI support for WWAN modems

[Manivannan Sadhasivam]

From: Loic Poulain <loic.poulain@linaro.org>

This is a generic MHI-over-PCI controller driver for MHI only devices
such as QCOM modems. For now it supports registering of Qualcomm SDX55
based PCIe modules. The MHI channels have been extracted from mhi
downstream driver.

This driver is for MHI-only devices which have all functionalities
exposed through MHI channels and accessed by the corresponding MHI
device drivers (no out-of-band communication).

Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Reviewed-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Hemant Kumar <hemantk@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
[mani: fixed up the Makefile rule]
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/Kconfig       |   9 +
 drivers/bus/mhi/Makefile      |   4 +
 drivers/bus/mhi/pci_generic.c | 345 ++++++++++++++++++++++++++++++++++
 3 files changed, 358 insertions(+)
 create mode 100644 drivers/bus/mhi/pci_generic.c

diff --git a/drivers/bus/mhi/Kconfig b/drivers/bus/mhi/Kconfig
index e841c1097fb4..da5cd0c9fc62 100644
--- a/drivers/bus/mhi/Kconfig
+++ b/drivers/bus/mhi/Kconfig
@@ -20,3 +20,12 @@ config MHI_BUS_DEBUG
 	  Enable debugfs support for use with the MHI transport. Allows
 	  reading and/or modifying some values within the MHI controller
 	  for debug and test purposes.
+
+config MHI_BUS_PCI_GENERIC
+	tristate "MHI PCI controller driver"
+	depends on MHI_BUS
+	depends on PCI
+	help
+	  This driver provides MHI PCI controller driver for devices such as
+	  Qualcomm SDX55 based PCIe modems.
+
diff --git a/drivers/bus/mhi/Makefile b/drivers/bus/mhi/Makefile
index 19e6443b72df..0a2d778d6fb4 100644
--- a/drivers/bus/mhi/Makefile
+++ b/drivers/bus/mhi/Makefile
@@ -1,2 +1,6 @@
 # core layer
 obj-y += core/
+
+obj-$(CONFIG_MHI_BUS_PCI_GENERIC) += mhi_pci_generic.o
+mhi_pci_generic-y += pci_generic.o
+
diff --git a/drivers/bus/mhi/pci_generic.c b/drivers/bus/mhi/pci_generic.c
new file mode 100644
index 000000000000..e3df838c3c80
--- /dev/null
+++ b/drivers/bus/mhi/pci_generic.c
@@ -0,0 +1,345 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * MHI PCI driver - MHI over PCI controller driver
+ *
+ * This module is a generic driver for registering MHI-over-PCI devices,
+ * such as PCIe QCOM modems.
+ *
+ * Copyright (C) 2020 Linaro Ltd <loic.poulain@linaro.org>
+ */
+
+#include <linux/device.h>
+#include <linux/mhi.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+
+#define MHI_PCI_DEFAULT_BAR_NUM 0
+
+/**
+ * struct mhi_pci_dev_info - MHI PCI device specific information
+ * @config: MHI controller configuration
+ * @name: name of the PCI module
+ * @fw: firmware path (if any)
+ * @edl: emergency download mode firmware path (if any)
+ * @bar_num: PCI base address register to use for MHI MMIO register space
+ * @dma_data_width: DMA transfer word size (32 or 64 bits)
+ */
+struct mhi_pci_dev_info {
+	const struct mhi_controller_config *config;
+	const char *name;
+	const char *fw;
+	const char *edl;
+	unsigned int bar_num;
+	unsigned int dma_data_width;
+};
+
+#define MHI_CHANNEL_CONFIG_UL(ch_num, ch_name, el_count, ev_ring) \
+	{						\
+		.num = ch_num,				\
+		.name = ch_name,			\
+		.num_elements = el_count,		\
+		.event_ring = ev_ring,			\
+		.dir = DMA_TO_DEVICE,			\
+		.ee_mask = BIT(MHI_EE_AMSS),		\
+		.pollcfg = 0,				\
+		.doorbell = MHI_DB_BRST_DISABLE,	\
+		.lpm_notify = false,			\
+		.offload_channel = false,		\
+		.doorbell_mode_switch = false,		\
+	}						\
+
+#define MHI_CHANNEL_CONFIG_DL(ch_num, ch_name, el_count, ev_ring) \
+	{						\
+		.num = ch_num,				\
+		.name = ch_name,			\
+		.num_elements = el_count,		\
+		.event_ring = ev_ring,			\
+		.dir = DMA_FROM_DEVICE,			\
+		.ee_mask = BIT(MHI_EE_AMSS),		\
+		.pollcfg = 0,				\
+		.doorbell = MHI_DB_BRST_DISABLE,	\
+		.lpm_notify = false,			\
+		.offload_channel = false,		\
+		.doorbell_mode_switch = false,		\
+	}
+
+#define MHI_EVENT_CONFIG_CTRL(ev_ring)		\
+	{					\
+		.num_elements = 64,		\
+		.irq_moderation_ms = 0,		\
+		.irq = (ev_ring) + 1,		\
+		.priority = 1,			\
+		.mode = MHI_DB_BRST_DISABLE,	\
+		.data_type = MHI_ER_CTRL,	\
+		.hardware_event = false,	\
+		.client_managed = false,	\
+		.offload_channel = false,	\
+	}
+
+#define MHI_EVENT_CONFIG_DATA(ev_ring)		\
+	{					\
+		.num_elements = 128,		\
+		.irq_moderation_ms = 5,		\
+		.irq = (ev_ring) + 1,		\
+		.priority = 1,			\
+		.mode = MHI_DB_BRST_DISABLE,	\
+		.data_type = MHI_ER_DATA,	\
+		.hardware_event = false,	\
+		.client_managed = false,	\
+		.offload_channel = false,	\
+	}
+
+#define MHI_EVENT_CONFIG_HW_DATA(ev_ring, ch_num) \
+	{					\
+		.num_elements = 128,		\
+		.irq_moderation_ms = 5,		\
+		.irq = (ev_ring) + 1,		\
+		.priority = 1,			\
+		.mode = MHI_DB_BRST_DISABLE,	\
+		.data_type = MHI_ER_DATA,	\
+		.hardware_event = true,		\
+		.client_managed = false,	\
+		.offload_channel = false,	\
+		.channel = ch_num,		\
+	}
+
+static const struct mhi_channel_config modem_qcom_v1_mhi_channels[] = {
+	MHI_CHANNEL_CONFIG_UL(12, "MBIM", 4, 0),
+	MHI_CHANNEL_CONFIG_DL(13, "MBIM", 4, 0),
+	MHI_CHANNEL_CONFIG_UL(14, "QMI", 4, 0),
+	MHI_CHANNEL_CONFIG_DL(15, "QMI", 4, 0),
+	MHI_CHANNEL_CONFIG_UL(20, "IPCR", 8, 0),
+	MHI_CHANNEL_CONFIG_DL(21, "IPCR", 8, 0),
+	MHI_CHANNEL_CONFIG_UL(100, "IP_HW0", 128, 1),
+	MHI_CHANNEL_CONFIG_DL(101, "IP_HW0", 128, 2),
+};
+
+static const struct mhi_event_config modem_qcom_v1_mhi_events[] = {
+	/* first ring is control+data ring */
+	MHI_EVENT_CONFIG_CTRL(0),
+	/* Hardware channels request dedicated hardware event rings */
+	MHI_EVENT_CONFIG_HW_DATA(1, 100),
+	MHI_EVENT_CONFIG_HW_DATA(2, 101)
+};
+
+static const struct mhi_controller_config modem_qcom_v1_mhiv_config = {
+	.max_channels = 128,
+	.timeout_ms = 5000,
+	.num_channels = ARRAY_SIZE(modem_qcom_v1_mhi_channels),
+	.ch_cfg = modem_qcom_v1_mhi_channels,
+	.num_events = ARRAY_SIZE(modem_qcom_v1_mhi_events),
+	.event_cfg = modem_qcom_v1_mhi_events,
+};
+
+static const struct mhi_pci_dev_info mhi_qcom_sdx55_info = {
+	.name = "qcom-sdx55m",
+	.fw = "qcom/sdx55m/sbl1.mbn",
+	.edl = "qcom/sdx55m/edl.mbn",
+	.config = &modem_qcom_v1_mhiv_config,
+	.bar_num = MHI_PCI_DEFAULT_BAR_NUM,
+	.dma_data_width = 32
+};
+
+static const struct pci_device_id mhi_pci_id_table[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_QCOM, 0x0306),
+		.driver_data = (kernel_ulong_t) &mhi_qcom_sdx55_info },
+	{  }
+};
+MODULE_DEVICE_TABLE(pci, mhi_pci_id_table);
+
+static int mhi_pci_read_reg(struct mhi_controller *mhi_cntrl,
+			    void __iomem *addr, u32 *out)
+{
+	*out = readl(addr);
+	return 0;
+}
+
+static void mhi_pci_write_reg(struct mhi_controller *mhi_cntrl,
+			      void __iomem *addr, u32 val)
+{
+	writel(val, addr);
+}
+
+static void mhi_pci_status_cb(struct mhi_controller *mhi_cntrl,
+			      enum mhi_callback cb)
+{
+	/* Nothing to do for now */
+}
+
+static int mhi_pci_claim(struct mhi_controller *mhi_cntrl,
+			 unsigned int bar_num, u64 dma_mask)
+{
+	struct pci_dev *pdev = to_pci_dev(mhi_cntrl->cntrl_dev);
+	int err;
+
+	err = pci_assign_resource(pdev, bar_num);
+	if (err)
+		return err;
+
+	err = pcim_enable_device(pdev);
+	if (err) {
+		dev_err(&pdev->dev, "failed to enable pci device: %d\n", err);
+		return err;
+	}
+
+	err = pcim_iomap_regions(pdev, 1 << bar_num, pci_name(pdev));
+	if (err) {
+		dev_err(&pdev->dev, "failed to map pci region: %d\n", err);
+		return err;
+	}
+	mhi_cntrl->regs = pcim_iomap_table(pdev)[bar_num];
+
+	err = pci_set_dma_mask(pdev, dma_mask);
+	if (err) {
+		dev_err(&pdev->dev, "Cannot set proper DMA mask\n");
+		return err;
+	}
+
+	err = pci_set_consistent_dma_mask(pdev, dma_mask);
+	if (err) {
+		dev_err(&pdev->dev, "set consistent dma mask failed\n");
+		return err;
+	}
+
+	pci_set_master(pdev);
+
+	return 0;
+}
+
+static int mhi_pci_get_irqs(struct mhi_controller *mhi_cntrl,
+			    const struct mhi_controller_config *mhi_cntrl_config)
+{
+	struct pci_dev *pdev = to_pci_dev(mhi_cntrl->cntrl_dev);
+	int nr_vectors, i;
+	int *irq;
+
+	/*
+	 * Alloc one MSI vector for BHI + one vector per event ring, ideally...
+	 * No explicit pci_free_irq_vectors required, done by pcim_release.
+	 */
+	mhi_cntrl->nr_irqs = 1 + mhi_cntrl_config->num_events;
+
+	nr_vectors = pci_alloc_irq_vectors(pdev, 1, mhi_cntrl->nr_irqs, PCI_IRQ_MSI);
+	if (nr_vectors < 0) {
+		dev_err(&pdev->dev, "Error allocating MSI vectors %d\n",
+			nr_vectors);
+		return nr_vectors;
+	}
+
+	if (nr_vectors < mhi_cntrl->nr_irqs) {
+		dev_warn(&pdev->dev, "Not enough MSI vectors (%d/%d), use shared MSI\n",
+			 nr_vectors, mhi_cntrl_config->num_events);
+	}
+
+	irq = devm_kcalloc(&pdev->dev, mhi_cntrl->nr_irqs, sizeof(int), GFP_KERNEL);
+	if (!irq)
+		return -ENOMEM;
+
+	for (i = 0; i < mhi_cntrl->nr_irqs; i++) {
+		int vector = i >= nr_vectors ? (nr_vectors - 1) : i;
+
+		irq[i] = pci_irq_vector(pdev, vector);
+	}
+
+	mhi_cntrl->irq = irq;
+
+	return 0;
+}
+
+static int mhi_pci_runtime_get(struct mhi_controller *mhi_cntrl)
+{
+	/* no PM for now */
+	return 0;
+}
+
+static void mhi_pci_runtime_put(struct mhi_controller *mhi_cntrl)
+{
+	/* no PM for now */
+}
+
+static int mhi_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+	const struct mhi_pci_dev_info *info = (struct mhi_pci_dev_info *) id->driver_data;
+	const struct mhi_controller_config *mhi_cntrl_config;
+	struct mhi_controller *mhi_cntrl;
+	int err;
+
+	dev_dbg(&pdev->dev, "MHI PCI device found: %s\n", info->name);
+
+	mhi_cntrl = mhi_alloc_controller();
+	if (!mhi_cntrl)
+		return -ENOMEM;
+
+	mhi_cntrl_config = info->config;
+	mhi_cntrl->cntrl_dev = &pdev->dev;
+	mhi_cntrl->iova_start = 0;
+	mhi_cntrl->iova_stop = DMA_BIT_MASK(info->dma_data_width);
+	mhi_cntrl->fw_image = info->fw;
+	mhi_cntrl->edl_image = info->edl;
+
+	mhi_cntrl->read_reg = mhi_pci_read_reg;
+	mhi_cntrl->write_reg = mhi_pci_write_reg;
+	mhi_cntrl->status_cb = mhi_pci_status_cb;
+	mhi_cntrl->runtime_get = mhi_pci_runtime_get;
+	mhi_cntrl->runtime_put = mhi_pci_runtime_put;
+
+	err = mhi_pci_claim(mhi_cntrl, info->bar_num, DMA_BIT_MASK(info->dma_data_width));
+	if (err)
+		goto err_release;
+
+	err = mhi_pci_get_irqs(mhi_cntrl, mhi_cntrl_config);
+	if (err)
+		goto err_release;
+
+	pci_set_drvdata(pdev, mhi_cntrl);
+
+	err = mhi_register_controller(mhi_cntrl, mhi_cntrl_config);
+	if (err)
+		goto err_release;
+
+	/* MHI bus does not power up the controller by default */
+	err = mhi_prepare_for_power_up(mhi_cntrl);
+	if (err) {
+		dev_err(&pdev->dev, "failed to prepare MHI controller\n");
+		goto err_unregister;
+	}
+
+	err = mhi_sync_power_up(mhi_cntrl);
+	if (err) {
+		dev_err(&pdev->dev, "failed to power up MHI controller\n");
+		goto err_unprepare;
+	}
+
+	return 0;
+
+err_unprepare:
+	mhi_unprepare_after_power_down(mhi_cntrl);
+err_unregister:
+	mhi_unregister_controller(mhi_cntrl);
+err_release:
+	mhi_free_controller(mhi_cntrl);
+
+	return err;
+}
+
+static void mhi_pci_remove(struct pci_dev *pdev)
+{
+	struct mhi_controller *mhi_cntrl = pci_get_drvdata(pdev);
+
+	mhi_power_down(mhi_cntrl, true);
+	mhi_unprepare_after_power_down(mhi_cntrl);
+	mhi_unregister_controller(mhi_cntrl);
+	mhi_free_controller(mhi_cntrl);
+}
+
+static struct pci_driver mhi_pci_driver = {
+	.name		= "mhi-pci-generic",
+	.id_table	= mhi_pci_id_table,
+	.probe		= mhi_pci_probe,
+	.remove		= mhi_pci_remove
+};
+module_pci_driver(mhi_pci_driver);
+
+MODULE_AUTHOR("Loic Poulain <loic.poulain@linaro.org>");
+MODULE_DESCRIPTION("Modem Host Interface (MHI) PCI controller driver");
+MODULE_LICENSE("GPL");
-- 
2.25.1

[PATCH 05/29] bus: mhi: core: Fix null pointer access when parsing MHI configuration

[Manivannan Sadhasivam]

From: Carl Yin <carl.yin@quectel.com>

Functions parse_ev_cfg() and parse_ch_cfg() access mhi_cntrl->mhi_dev
before it is set in function mhi_register_controller(),
use cntrl_dev instead of mhi_dev.

Fixes: 0cbf260820fa ("bus: mhi: core: Add support for registering MHI controllers")
Signed-off-by: Carl Yin <carl.yin@quectel.com>
Reviewed-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Hemant Kumar <hemantk@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/init.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/bus/mhi/core/init.c b/drivers/bus/mhi/core/init.c
index 0a09f8215057..8cefa359fccd 100644
--- a/drivers/bus/mhi/core/init.c
+++ b/drivers/bus/mhi/core/init.c
@@ -610,7 +610,7 @@ static int parse_ev_cfg(struct mhi_controller *mhi_cntrl,
 {
 	struct mhi_event *mhi_event;
 	const struct mhi_event_config *event_cfg;
-	struct device *dev = &mhi_cntrl->mhi_dev->dev;
+	struct device *dev = mhi_cntrl->cntrl_dev;
 	int i, num;
 
 	num = config->num_events;
@@ -692,7 +692,7 @@ static int parse_ch_cfg(struct mhi_controller *mhi_cntrl,
 			const struct mhi_controller_config *config)
 {
 	const struct mhi_channel_config *ch_cfg;
-	struct device *dev = &mhi_cntrl->mhi_dev->dev;
+	struct device *dev = mhi_cntrl->cntrl_dev;
 	int i;
 	u32 chan;
 
-- 
2.25.1

[PATCH 06/29] bus: mhi: Fix channel close issue on driver remove

[Manivannan Sadhasivam]

From: Loic Poulain <loic.poulain@linaro.org>

Some MHI device drivers need to stop the channels in their driver
remove callback (e.g. module unloading), but the unprepare function
is aborted because MHI core moved the channels to suspended state
prior calling driver remove callback. This prevents the driver to
send a proper MHI RESET CHAN command to the device. Device is then
unaware of the stopped state of these channels.

This causes issue when driver tries to start the channels again (e.g.
module is reloaded), since device considers channels as already
started (inconsistent state).

Fix this by allowing channel reset when channel is suspended.

Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Reviewed-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/main.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/bus/mhi/core/main.c b/drivers/bus/mhi/core/main.c
index 2cff5ddff225..cea5eab83f48 100644
--- a/drivers/bus/mhi/core/main.c
+++ b/drivers/bus/mhi/core/main.c
@@ -1224,7 +1224,8 @@ static void __mhi_unprepare_channel(struct mhi_controller *mhi_cntrl,
 	/* no more processing events for this channel */
 	mutex_lock(&mhi_chan->mutex);
 	write_lock_irq(&mhi_chan->lock);
-	if (mhi_chan->ch_state != MHI_CH_STATE_ENABLED) {
+	if (mhi_chan->ch_state != MHI_CH_STATE_ENABLED &&
+	    mhi_chan->ch_state != MHI_CH_STATE_SUSPENDED) {
 		write_unlock_irq(&mhi_chan->lock);
 		mutex_unlock(&mhi_chan->mutex);
 		return;
-- 
2.25.1

[PATCH 07/29] bus: mhi: core: Remove unnecessary counter from mhi_firmware_copy()

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

There is an extra 'i' counter in the mhi_firmware_copy() function
which is unused. Remove it to clean-up code and reduce stack
space as well as improve efficiency of the function.

Fixes: cd457afb1667 ("bus: mhi: core: Add support for downloading firmware over BHIe")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/boot.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/drivers/bus/mhi/core/boot.c b/drivers/bus/mhi/core/boot.c
index 24422f5c3d80..6b6fd9668c3b 100644
--- a/drivers/bus/mhi/core/boot.c
+++ b/drivers/bus/mhi/core/boot.c
@@ -365,7 +365,6 @@ static void mhi_firmware_copy(struct mhi_controller *mhi_cntrl,
 	size_t remainder = firmware->size;
 	size_t to_cpy;
 	const u8 *buf = firmware->data;
-	int i = 0;
 	struct mhi_buf *mhi_buf = img_info->mhi_buf;
 	struct bhi_vec_entry *bhi_vec = img_info->bhi_vec;
 
@@ -377,7 +376,6 @@ static void mhi_firmware_copy(struct mhi_controller *mhi_cntrl,
 
 		buf += to_cpy;
 		remainder -= to_cpy;
-		i++;
 		bhi_vec++;
 		mhi_buf++;
 	}
-- 
2.25.1

[PATCH 08/29] bus: mhi: core: Add missing EXPORT_SYMBOL for mhi_get_mhi_state()

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

Add missing EXPORT_SYMBOL_GPL() declaration for mhi_get_mhi_state()
API.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/main.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/bus/mhi/core/main.c b/drivers/bus/mhi/core/main.c
index cea5eab83f48..6ecaacaa8b54 100644
--- a/drivers/bus/mhi/core/main.c
+++ b/drivers/bus/mhi/core/main.c
@@ -132,6 +132,7 @@ enum mhi_state mhi_get_mhi_state(struct mhi_controller *mhi_cntrl)
 				     MHISTATUS_MHISTATE_SHIFT, &state);
 	return ret ? MHI_STATE_MAX : state;
 }
+EXPORT_SYMBOL_GPL(mhi_get_mhi_state);
 
 int mhi_map_single_no_bb(struct mhi_controller *mhi_cntrl,
 			 struct mhi_buf_info *buf_info)
-- 
2.25.1

[PATCH 09/29] bus: mhi: core: Expose mhi_get_exec_env() API for controllers

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

The mhi_get_exec_env() APIs can be used by the controller drivers
to query the execution environment of the MHI device. Expose it
so it can be used in some scenarios to determine behavior of
controllers.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/internal.h | 1 -
 drivers/bus/mhi/core/main.c     | 1 +
 include/linux/mhi.h             | 6 ++++++
 3 files changed, 7 insertions(+), 1 deletion(-)

diff --git a/drivers/bus/mhi/core/internal.h b/drivers/bus/mhi/core/internal.h
index 78e4e84d6743..d8af8a702493 100644
--- a/drivers/bus/mhi/core/internal.h
+++ b/drivers/bus/mhi/core/internal.h
@@ -609,7 +609,6 @@ enum mhi_pm_state __must_check mhi_tryset_pm_state(
 					struct mhi_controller *mhi_cntrl,
 					enum mhi_pm_state state);
 const char *to_mhi_pm_state_str(enum mhi_pm_state state);
-enum mhi_ee_type mhi_get_exec_env(struct mhi_controller *mhi_cntrl);
 int mhi_queue_state_transition(struct mhi_controller *mhi_cntrl,
 			       enum dev_st_transition state);
 void mhi_pm_st_worker(struct work_struct *work);
diff --git a/drivers/bus/mhi/core/main.c b/drivers/bus/mhi/core/main.c
index 6ecaacaa8b54..f953e2a6d58a 100644
--- a/drivers/bus/mhi/core/main.c
+++ b/drivers/bus/mhi/core/main.c
@@ -123,6 +123,7 @@ enum mhi_ee_type mhi_get_exec_env(struct mhi_controller *mhi_cntrl)
 
 	return (ret) ? MHI_EE_MAX : exec;
 }
+EXPORT_SYMBOL_GPL(mhi_get_exec_env);
 
 enum mhi_state mhi_get_mhi_state(struct mhi_controller *mhi_cntrl)
 {
diff --git a/include/linux/mhi.h b/include/linux/mhi.h
index d4841e5a5f45..9225d5551d69 100644
--- a/include/linux/mhi.h
+++ b/include/linux/mhi.h
@@ -658,6 +658,12 @@ int mhi_download_rddm_img(struct mhi_controller *mhi_cntrl, bool in_panic);
  */
 int mhi_force_rddm_mode(struct mhi_controller *mhi_cntrl);
 
+/**
+ * mhi_get_exec_env - Get BHI execution environment of the device
+ * @mhi_cntrl: MHI controller
+ */
+enum mhi_ee_type mhi_get_exec_env(struct mhi_controller *mhi_cntrl);
+
 /**
  * mhi_get_mhi_state - Get MHI state of the device
  * @mhi_cntrl: MHI controller
-- 
2.25.1

[PATCH 10/29] bus: mhi: core: Remove unused mhi_fw_load_worker() declaration

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

The mhi_fw_load_worker() function no longer exists. Remove its
declaration as part of code clean-up.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/internal.h | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/bus/mhi/core/internal.h b/drivers/bus/mhi/core/internal.h
index d8af8a702493..2df8de5432f9 100644
--- a/drivers/bus/mhi/core/internal.h
+++ b/drivers/bus/mhi/core/internal.h
@@ -613,7 +613,6 @@ int mhi_queue_state_transition(struct mhi_controller *mhi_cntrl,
 			       enum dev_st_transition state);
 void mhi_pm_st_worker(struct work_struct *work);
 void mhi_pm_sys_err_handler(struct mhi_controller *mhi_cntrl);
-void mhi_fw_load_worker(struct work_struct *work);
 int mhi_ready_state_transition(struct mhi_controller *mhi_cntrl);
 int mhi_pm_m0_transition(struct mhi_controller *mhi_cntrl);
 void mhi_pm_m1_transition(struct mhi_controller *mhi_cntrl);
-- 
2.25.1

[PATCH 11/29] bus: mhi: core: Rename RDDM download function to use proper words

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

mhi_download_rddm_img() uses a shorter version of the word image.
Expand it and rename the function to mhi_download_rddm_image().

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/boot.c | 4 ++--
 include/linux/mhi.h         | 6 +++---
 2 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/drivers/bus/mhi/core/boot.c b/drivers/bus/mhi/core/boot.c
index 6b6fd9668c3b..16244cc8fbe7 100644
--- a/drivers/bus/mhi/core/boot.c
+++ b/drivers/bus/mhi/core/boot.c
@@ -147,7 +147,7 @@ static int __mhi_download_rddm_in_panic(struct mhi_controller *mhi_cntrl)
 }
 
 /* Download RDDM image from device */
-int mhi_download_rddm_img(struct mhi_controller *mhi_cntrl, bool in_panic)
+int mhi_download_rddm_image(struct mhi_controller *mhi_cntrl, bool in_panic)
 {
 	void __iomem *base = mhi_cntrl->bhie;
 	struct device *dev = &mhi_cntrl->mhi_dev->dev;
@@ -169,7 +169,7 @@ int mhi_download_rddm_img(struct mhi_controller *mhi_cntrl, bool in_panic)
 
 	return (rx_status == BHIE_RXVECSTATUS_STATUS_XFER_COMPL) ? 0 : -EIO;
 }
-EXPORT_SYMBOL_GPL(mhi_download_rddm_img);
+EXPORT_SYMBOL_GPL(mhi_download_rddm_image);
 
 static int mhi_fw_load_amss(struct mhi_controller *mhi_cntrl,
 			    const struct mhi_buf *mhi_buf)
diff --git a/include/linux/mhi.h b/include/linux/mhi.h
index 9225d5551d69..52b3c60bf9bb 100644
--- a/include/linux/mhi.h
+++ b/include/linux/mhi.h
@@ -645,12 +645,12 @@ int mhi_pm_suspend(struct mhi_controller *mhi_cntrl);
 int mhi_pm_resume(struct mhi_controller *mhi_cntrl);
 
 /**
- * mhi_download_rddm_img - Download ramdump image from device for
- *                         debugging purpose.
+ * mhi_download_rddm_image - Download ramdump image from device for
+ *                           debugging purpose.
  * @mhi_cntrl: MHI controller
  * @in_panic: Download rddm image during kernel panic
  */
-int mhi_download_rddm_img(struct mhi_controller *mhi_cntrl, bool in_panic);
+int mhi_download_rddm_image(struct mhi_controller *mhi_cntrl, bool in_panic);
 
 /**
  * mhi_force_rddm_mode - Force device into rddm mode
-- 
2.25.1

[PATCH 12/29] bus: mhi: core: Skip RDDM download for unknown execution environment

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

If MHI is unable to determine the execution environment during
the panic path, host must skip the RDDM download. This can happen
if the BHI offset read or the BHI_EXECENV register read fails
indicating that the underlying transport is unresponsive. Hence,
there is no need to trigger an RDDM using SYSERR or request an
SOC reset.

Suggested-by: Hemant Kumar <hemantk@codeaurora.org>
Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/boot.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/drivers/bus/mhi/core/boot.c b/drivers/bus/mhi/core/boot.c
index 16244cc8fbe7..6f0cfb9922b4 100644
--- a/drivers/bus/mhi/core/boot.c
+++ b/drivers/bus/mhi/core/boot.c
@@ -92,6 +92,9 @@ static int __mhi_download_rddm_in_panic(struct mhi_controller *mhi_cntrl)
 	 * image download completion.
 	 */
 	ee = mhi_get_exec_env(mhi_cntrl);
+	if (ee == MHI_EE_MAX)
+		goto error_exit_rddm;
+
 	if (ee != MHI_EE_RDDM) {
 		dev_dbg(dev, "Trigger device into RDDM mode using SYS ERR\n");
 		mhi_set_mhi_state(mhi_cntrl, MHI_STATE_SYS_ERR);
@@ -139,10 +142,12 @@ static int __mhi_download_rddm_in_panic(struct mhi_controller *mhi_cntrl)
 	ee = mhi_get_exec_env(mhi_cntrl);
 	ret = mhi_read_reg(mhi_cntrl, base, BHIE_RXVECSTATUS_OFFS, &rx_status);
 
-	dev_err(dev, "Did not complete RDDM transfer\n");
-	dev_err(dev, "Current EE: %s\n", TO_MHI_EXEC_STR(ee));
 	dev_err(dev, "RXVEC_STATUS: 0x%x\n", rx_status);
 
+error_exit_rddm:
+	dev_err(dev, "RDDM transfer failed. Current EE: %s\n",
+		TO_MHI_EXEC_STR(ee));
+
 	return -EIO;
 }
 
-- 
2.25.1

[PATCH 13/29] bus: mhi: core: Use appropriate names for firmware load functions

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

mhi_fw_load_sbl() function is currently used to transfer SBL or EDL
images over BHI (Boot Host Interface). Same goes with mhi_fw_load_amss()
which uses BHIe. However, the contents of these functions do not
indicate support for a specific set of images. Since these can be used
for any image download over BHI or BHIe, rename them based on the
protocol used.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/boot.c | 19 ++++++++++---------
 1 file changed, 10 insertions(+), 9 deletions(-)

diff --git a/drivers/bus/mhi/core/boot.c b/drivers/bus/mhi/core/boot.c
index 6f0cfb9922b4..2d7752cd9ccc 100644
--- a/drivers/bus/mhi/core/boot.c
+++ b/drivers/bus/mhi/core/boot.c
@@ -176,7 +176,7 @@ int mhi_download_rddm_image(struct mhi_controller *mhi_cntrl, bool in_panic)
 }
 EXPORT_SYMBOL_GPL(mhi_download_rddm_image);
 
-static int mhi_fw_load_amss(struct mhi_controller *mhi_cntrl,
+static int mhi_fw_load_bhie(struct mhi_controller *mhi_cntrl,
 			    const struct mhi_buf *mhi_buf)
 {
 	void __iomem *base = mhi_cntrl->bhie;
@@ -192,7 +192,7 @@ static int mhi_fw_load_amss(struct mhi_controller *mhi_cntrl,
 	}
 
 	sequence_id = MHI_RANDOM_U32_NONZERO(BHIE_TXVECSTATUS_SEQNUM_BMSK);
-	dev_dbg(dev, "Starting AMSS download via BHIe. Sequence ID:%u\n",
+	dev_dbg(dev, "Starting image download via BHIe. Sequence ID: %u\n",
 		sequence_id);
 	mhi_write_reg(mhi_cntrl, base, BHIE_TXVECADDR_HIGH_OFFS,
 		      upper_32_bits(mhi_buf->dma_addr));
@@ -223,7 +223,7 @@ static int mhi_fw_load_amss(struct mhi_controller *mhi_cntrl,
 	return (!ret) ? -ETIMEDOUT : 0;
 }
 
-static int mhi_fw_load_sbl(struct mhi_controller *mhi_cntrl,
+static int mhi_fw_load_bhi(struct mhi_controller *mhi_cntrl,
 			   dma_addr_t dma_addr,
 			   size_t size)
 {
@@ -250,7 +250,7 @@ static int mhi_fw_load_sbl(struct mhi_controller *mhi_cntrl,
 	}
 
 	session_id = MHI_RANDOM_U32_NONZERO(BHI_TXDB_SEQNUM_BMSK);
-	dev_dbg(dev, "Starting SBL download via BHI. Session ID:%u\n",
+	dev_dbg(dev, "Starting image download via BHI. Session ID: %u\n",
 		session_id);
 	mhi_write_reg(mhi_cntrl, base, BHI_STATUS, 0);
 	mhi_write_reg(mhi_cntrl, base, BHI_IMGADDR_HIGH,
@@ -449,9 +449,9 @@ void mhi_fw_load_handler(struct mhi_controller *mhi_cntrl)
 		return;
 	}
 
-	/* Download SBL image */
+	/* Download image using BHI */
 	memcpy(buf, firmware->data, size);
-	ret = mhi_fw_load_sbl(mhi_cntrl, dma_addr, size);
+	ret = mhi_fw_load_bhi(mhi_cntrl, dma_addr, size);
 	mhi_free_coherent(mhi_cntrl, size, buf, dma_addr);
 
 	if (!mhi_cntrl->fbc_download || ret || mhi_cntrl->ee == MHI_EE_EDL)
@@ -459,7 +459,7 @@ void mhi_fw_load_handler(struct mhi_controller *mhi_cntrl)
 
 	/* Error or in EDL mode, we're done */
 	if (ret) {
-		dev_err(dev, "MHI did not load SBL, ret:%d\n", ret);
+		dev_err(dev, "MHI did not load image over BHI, ret: %d\n", ret);
 		return;
 	}
 
@@ -509,11 +509,12 @@ void mhi_fw_load_handler(struct mhi_controller *mhi_cntrl)
 
 	/* Start full firmware image download */
 	image_info = mhi_cntrl->fbc_image;
-	ret = mhi_fw_load_amss(mhi_cntrl,
+	ret = mhi_fw_load_bhie(mhi_cntrl,
 			       /* Vector table is the last entry */
 			       &image_info->mhi_buf[image_info->entries - 1]);
 	if (ret)
-		dev_err(dev, "MHI did not load AMSS, ret:%d\n", ret);
+		dev_err(dev, "MHI did not load image over BHIe, ret: %d\n",
+			ret);
 
 	release_firmware(firmware);
 
-- 
2.25.1

[PATCH 14/29] bus: mhi: core: Move to using high priority workqueue

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

MHI work is currently scheduled on the global/system workqueue and can
encounter delays on a stressed system. To avoid those unforeseen
delays which can hamper bootup or shutdown times, use a dedicated high
priority workqueue instead of the global/system workqueue.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/init.c | 9 +++++++++
 drivers/bus/mhi/core/pm.c   | 2 +-
 include/linux/mhi.h         | 2 ++
 3 files changed, 12 insertions(+), 1 deletion(-)

diff --git a/drivers/bus/mhi/core/init.c b/drivers/bus/mhi/core/init.c
index 8cefa359fccd..877e40c86801 100644
--- a/drivers/bus/mhi/core/init.c
+++ b/drivers/bus/mhi/core/init.c
@@ -880,6 +880,13 @@ int mhi_register_controller(struct mhi_controller *mhi_cntrl,
 	INIT_WORK(&mhi_cntrl->st_worker, mhi_pm_st_worker);
 	init_waitqueue_head(&mhi_cntrl->state_event);
 
+	mhi_cntrl->hiprio_wq = alloc_ordered_workqueue
+				("mhi_hiprio_wq", WQ_MEM_RECLAIM | WQ_HIGHPRI);
+	if (!mhi_cntrl->hiprio_wq) {
+		dev_err(mhi_cntrl->cntrl_dev, "Failed to allocate workqueue\n");
+		goto error_alloc_cmd;
+	}
+
 	mhi_cmd = mhi_cntrl->mhi_cmd;
 	for (i = 0; i < NR_OF_CMD_RINGS; i++, mhi_cmd++)
 		spin_lock_init(&mhi_cmd->lock);
@@ -969,6 +976,7 @@ int mhi_register_controller(struct mhi_controller *mhi_cntrl,
 error_alloc_cmd:
 	vfree(mhi_cntrl->mhi_chan);
 	kfree(mhi_cntrl->mhi_event);
+	destroy_workqueue(mhi_cntrl->hiprio_wq);
 
 	return ret;
 }
@@ -982,6 +990,7 @@ void mhi_unregister_controller(struct mhi_controller *mhi_cntrl)
 
 	mhi_destroy_debugfs(mhi_cntrl);
 
+	destroy_workqueue(mhi_cntrl->hiprio_wq);
 	kfree(mhi_cntrl->mhi_cmd);
 	kfree(mhi_cntrl->mhi_event);
 
diff --git a/drivers/bus/mhi/core/pm.c b/drivers/bus/mhi/core/pm.c
index 3de7b1639ec6..805b6fa748f0 100644
--- a/drivers/bus/mhi/core/pm.c
+++ b/drivers/bus/mhi/core/pm.c
@@ -597,7 +597,7 @@ int mhi_queue_state_transition(struct mhi_controller *mhi_cntrl,
 	list_add_tail(&item->node, &mhi_cntrl->transition_list);
 	spin_unlock_irqrestore(&mhi_cntrl->transition_lock, flags);
 
-	schedule_work(&mhi_cntrl->st_worker);
+	queue_work(mhi_cntrl->hiprio_wq, &mhi_cntrl->st_worker);
 
 	return 0;
 }
diff --git a/include/linux/mhi.h b/include/linux/mhi.h
index 52b3c60bf9bb..1ed5f2aa224b 100644
--- a/include/linux/mhi.h
+++ b/include/linux/mhi.h
@@ -337,6 +337,7 @@ struct mhi_controller_config {
  * @wlock: Lock for protecting device wakeup
  * @mhi_link_info: Device bandwidth info
  * @st_worker: State transition worker
+ * @hiprio_wq: High priority workqueue for MHI work such as state transitions
  * @state_event: State change event
  * @status_cb: CB function to notify power states of the device (required)
  * @wake_get: CB function to assert device wake (optional)
@@ -419,6 +420,7 @@ struct mhi_controller {
 	spinlock_t wlock;
 	struct mhi_link_info mhi_link_info;
 	struct work_struct st_worker;
+	struct workqueue_struct *hiprio_wq;
 	wait_queue_head_t state_event;
 
 	void (*status_cb)(struct mhi_controller *mhi_cntrl,
-- 
2.25.1

[PATCH 15/29] bus: mhi: core: Skip device wake in error or shutdown states

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

MHI client drivers can request a device wake even if the device
may be in an error state or undergoing a shutdown. To prevent
unnecessary device wake processing, check for the device state
and bail out early so that the clients are made aware of the
device state sooner.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/pm.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/bus/mhi/core/pm.c b/drivers/bus/mhi/core/pm.c
index 805b6fa748f0..029919647002 100644
--- a/drivers/bus/mhi/core/pm.c
+++ b/drivers/bus/mhi/core/pm.c
@@ -827,6 +827,10 @@ int __mhi_device_get_sync(struct mhi_controller *mhi_cntrl)
 
 	/* Wake up the device */
 	read_lock_bh(&mhi_cntrl->pm_lock);
+	if (MHI_PM_IN_ERROR_STATE(mhi_cntrl->pm_state)) {
+		read_unlock_bh(&mhi_cntrl->pm_lock);
+		return -EIO;
+	}
 	mhi_cntrl->wake_get(mhi_cntrl, true);
 	if (MHI_PM_IN_SUSPEND_STATE(mhi_cntrl->pm_state))
 		mhi_trigger_resume(mhi_cntrl);
-- 
2.25.1

[PATCH 16/29] bus: mhi: core: Move to SYS_ERROR regardless of RDDM capability

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

In some cases, the entry of device to RDDM execution environment
can occur after a significant amount of time has elapsed and a
SYS_ERROR state change event has already arrived. This can result
in scenarios where MHI controller and client drivers are unaware
of the error state of the device. Remove the check for rddm_image
when processing the SYS_ERROR state change as it is present in
mhi_pm_sys_err_handler() already and prevent further activity
until the expected RDDM execution environment change occurs or
the controller driver decides further action.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/main.c | 5 -----
 1 file changed, 5 deletions(-)

diff --git a/drivers/bus/mhi/core/main.c b/drivers/bus/mhi/core/main.c
index f953e2a6d58a..91f8b8d05a62 100644
--- a/drivers/bus/mhi/core/main.c
+++ b/drivers/bus/mhi/core/main.c
@@ -737,11 +737,6 @@ int mhi_process_ctrl_ev_ring(struct mhi_controller *mhi_cntrl,
 			{
 				enum mhi_pm_state new_state;
 
-				/* skip SYS_ERROR handling if RDDM supported */
-				if (mhi_cntrl->ee == MHI_EE_RDDM ||
-				    mhi_cntrl->rddm_image)
-					break;
-
 				dev_dbg(dev, "System error detected\n");
 				write_lock_irq(&mhi_cntrl->pm_lock);
 				new_state = mhi_tryset_pm_state(mhi_cntrl,
-- 
2.25.1

[PATCH 17/29] bus: mhi: core: Prevent sending multiple RDDM entry callbacks

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

If an mhi_power_down() is initiated after the device has entered
RDDM and a status callback was provided for it, it is possible
that another BHI interrupt fires while waiting for the MHI
RESET to be cleared. If that happens, MHI host would have moved
a "disabled" execution environment and the check to allow sending
an RDDM status callback will pass when it is should not. Add a
check to see if MHI is in an active state before proceeding.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/main.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/bus/mhi/core/main.c b/drivers/bus/mhi/core/main.c
index 91f8b8d05a62..4eb93d8bea1d 100644
--- a/drivers/bus/mhi/core/main.c
+++ b/drivers/bus/mhi/core/main.c
@@ -401,6 +401,10 @@ irqreturn_t mhi_intvec_threaded_handler(int irq_number, void *priv)
 
 	 /* If device supports RDDM don't bother processing SYS error */
 	if (mhi_cntrl->rddm_image) {
+		/* host may be performing a device power down already */
+		if (!mhi_is_active(mhi_cntrl))
+			goto exit_intvec;
+
 		if (mhi_cntrl->ee == MHI_EE_RDDM && mhi_cntrl->ee != ee) {
 			mhi_cntrl->status_cb(mhi_cntrl, MHI_CB_EE_RDDM);
 			wake_up_all(&mhi_cntrl->state_event);
-- 
2.25.1

[PATCH 18/29] bus: mhi: core: Move to an error state on any firmware load failure

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

Move MHI to a firmware download error state for a failure to find
the firmware files or to load SBL or EBL image using BHI/BHIe. This
helps detect an error state sooner and shortens the wait for a
synchronous power up timeout.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/boot.c | 35 ++++++++++++++++++++---------------
 1 file changed, 20 insertions(+), 15 deletions(-)

diff --git a/drivers/bus/mhi/core/boot.c b/drivers/bus/mhi/core/boot.c
index 2d7752cd9ccc..7b57bb9a3080 100644
--- a/drivers/bus/mhi/core/boot.c
+++ b/drivers/bus/mhi/core/boot.c
@@ -428,13 +428,13 @@ void mhi_fw_load_handler(struct mhi_controller *mhi_cntrl)
 						     !mhi_cntrl->seg_len))) {
 		dev_err(dev,
 			"No firmware image defined or !sbl_size || !seg_len\n");
-		return;
+		goto error_fw_load;
 	}
 
 	ret = request_firmware(&firmware, fw_name, dev);
 	if (ret) {
 		dev_err(dev, "Error loading firmware: %d\n", ret);
-		return;
+		goto error_fw_load;
 	}
 
 	size = (mhi_cntrl->fbc_download) ? mhi_cntrl->sbl_size : firmware->size;
@@ -446,7 +446,7 @@ void mhi_fw_load_handler(struct mhi_controller *mhi_cntrl)
 	buf = mhi_alloc_coherent(mhi_cntrl, size, &dma_addr, GFP_KERNEL);
 	if (!buf) {
 		release_firmware(firmware);
-		return;
+		goto error_fw_load;
 	}
 
 	/* Download image using BHI */
@@ -454,17 +454,17 @@ void mhi_fw_load_handler(struct mhi_controller *mhi_cntrl)
 	ret = mhi_fw_load_bhi(mhi_cntrl, dma_addr, size);
 	mhi_free_coherent(mhi_cntrl, size, buf, dma_addr);
 
-	if (!mhi_cntrl->fbc_download || ret || mhi_cntrl->ee == MHI_EE_EDL)
-		release_firmware(firmware);
-
 	/* Error or in EDL mode, we're done */
 	if (ret) {
 		dev_err(dev, "MHI did not load image over BHI, ret: %d\n", ret);
-		return;
+		release_firmware(firmware);
+		goto error_fw_load;
 	}
 
-	if (mhi_cntrl->ee == MHI_EE_EDL)
+	if (mhi_cntrl->ee == MHI_EE_EDL) {
+		release_firmware(firmware);
 		return;
+	}
 
 	write_lock_irq(&mhi_cntrl->pm_lock);
 	mhi_cntrl->dev_state = MHI_STATE_RESET;
@@ -477,13 +477,17 @@ void mhi_fw_load_handler(struct mhi_controller *mhi_cntrl)
 	if (mhi_cntrl->fbc_download) {
 		ret = mhi_alloc_bhie_table(mhi_cntrl, &mhi_cntrl->fbc_image,
 					   firmware->size);
-		if (ret)
-			goto error_alloc_fw_table;
+		if (ret) {
+			release_firmware(firmware);
+			goto error_fw_load;
+		}
 
 		/* Load the firmware into BHIE vec table */
 		mhi_firmware_copy(mhi_cntrl, firmware, mhi_cntrl->fbc_image);
 	}
 
+	release_firmware(firmware);
+
 fw_load_ee_pthru:
 	/* Transitioning into MHI RESET->READY state */
 	ret = mhi_ready_state_transition(mhi_cntrl);
@@ -512,11 +516,11 @@ void mhi_fw_load_handler(struct mhi_controller *mhi_cntrl)
 	ret = mhi_fw_load_bhie(mhi_cntrl,
 			       /* Vector table is the last entry */
 			       &image_info->mhi_buf[image_info->entries - 1]);
-	if (ret)
+	if (ret) {
 		dev_err(dev, "MHI did not load image over BHIe, ret: %d\n",
 			ret);
-
-	release_firmware(firmware);
+		goto error_fw_load;
+	}
 
 	return;
 
@@ -524,6 +528,7 @@ void mhi_fw_load_handler(struct mhi_controller *mhi_cntrl)
 	mhi_free_bhie_table(mhi_cntrl, mhi_cntrl->fbc_image);
 	mhi_cntrl->fbc_image = NULL;
 
-error_alloc_fw_table:
-	release_firmware(firmware);
+error_fw_load:
+	mhi_cntrl->pm_state = MHI_PM_FW_DL_ERR;
+	wake_up_all(&mhi_cntrl->state_event);
 }
-- 
2.25.1

[PATCH 19/29] bus: mhi: core: Use appropriate label in firmware load handler API

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

Correct the "error_read" label to say "error_ready_state" as that
is the appropriate usage of the label.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/boot.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/bus/mhi/core/boot.c b/drivers/bus/mhi/core/boot.c
index 7b57bb9a3080..c2546bf229fb 100644
--- a/drivers/bus/mhi/core/boot.c
+++ b/drivers/bus/mhi/core/boot.c
@@ -497,7 +497,7 @@ void mhi_fw_load_handler(struct mhi_controller *mhi_cntrl)
 
 	if (ret) {
 		dev_err(dev, "MHI did not enter READY state\n");
-		goto error_read;
+		goto error_ready_state;
 	}
 
 	/* Wait for the SBL event */
@@ -508,7 +508,7 @@ void mhi_fw_load_handler(struct mhi_controller *mhi_cntrl)
 
 	if (!ret || MHI_PM_IN_ERROR_STATE(mhi_cntrl->pm_state)) {
 		dev_err(dev, "MHI did not enter SBL\n");
-		goto error_read;
+		goto error_ready_state;
 	}
 
 	/* Start full firmware image download */
@@ -524,7 +524,7 @@ void mhi_fw_load_handler(struct mhi_controller *mhi_cntrl)
 
 	return;
 
-error_read:
+error_ready_state:
 	mhi_free_bhie_table(mhi_cntrl, mhi_cntrl->fbc_image);
 	mhi_cntrl->fbc_image = NULL;
 
-- 
2.25.1

[PATCH 20/29] bus: mhi: core: Move to an error state on mission mode failure

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

If the host receives a mission mode event and by the time it can get
to processing it, the register accesses fail implying a connectivity
error, MHI should move to an error state. This helps avoid longer wait
times from a synchronous power up perspective and accurately reflects
the MHI execution environment and power management states.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/pm.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/bus/mhi/core/pm.c b/drivers/bus/mhi/core/pm.c
index 029919647002..06adea2580d2 100644
--- a/drivers/bus/mhi/core/pm.c
+++ b/drivers/bus/mhi/core/pm.c
@@ -383,10 +383,14 @@ static int mhi_pm_mission_mode_transition(struct mhi_controller *mhi_cntrl)
 	write_lock_irq(&mhi_cntrl->pm_lock);
 	if (MHI_REG_ACCESS_VALID(mhi_cntrl->pm_state))
 		mhi_cntrl->ee = mhi_get_exec_env(mhi_cntrl);
-	write_unlock_irq(&mhi_cntrl->pm_lock);
 
-	if (!MHI_IN_MISSION_MODE(mhi_cntrl->ee))
+	if (!MHI_IN_MISSION_MODE(mhi_cntrl->ee)) {
+		mhi_cntrl->pm_state = MHI_PM_LD_ERR_FATAL_DETECT;
+		write_unlock_irq(&mhi_cntrl->pm_lock);
+		wake_up_all(&mhi_cntrl->state_event);
 		return -EIO;
+	}
+	write_unlock_irq(&mhi_cntrl->pm_lock);
 
 	wake_up_all(&mhi_cntrl->state_event);
 
-- 
2.25.1

[PATCH 21/29] bus: mhi: core: Check for IRQ availability during registration

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

Current design allows a controller to register with MHI successfully
without the need to have any IRQs available for use. If no IRQs are
available, power up requests to MHI can fail after a successful
registration with MHI. Improve the design by checking for the number
of IRQs available sooner within the mhi_regsiter_controller() API as
it is required to be specified by the controller.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/init.c | 2 +-
 drivers/bus/mhi/core/pm.c   | 3 ---
 2 files changed, 1 insertion(+), 4 deletions(-)

diff --git a/drivers/bus/mhi/core/init.c b/drivers/bus/mhi/core/init.c
index 877e40c86801..2534f1c9c153 100644
--- a/drivers/bus/mhi/core/init.c
+++ b/drivers/bus/mhi/core/init.c
@@ -858,7 +858,7 @@ int mhi_register_controller(struct mhi_controller *mhi_cntrl,
 
 	if (!mhi_cntrl->runtime_get || !mhi_cntrl->runtime_put ||
 	    !mhi_cntrl->status_cb || !mhi_cntrl->read_reg ||
-	    !mhi_cntrl->write_reg)
+	    !mhi_cntrl->write_reg || !mhi_cntrl->nr_irqs)
 		return -EINVAL;
 
 	ret = parse_config(mhi_cntrl, config);
diff --git a/drivers/bus/mhi/core/pm.c b/drivers/bus/mhi/core/pm.c
index 06adea2580d2..1d04e401b67f 100644
--- a/drivers/bus/mhi/core/pm.c
+++ b/drivers/bus/mhi/core/pm.c
@@ -926,9 +926,6 @@ int mhi_async_power_up(struct mhi_controller *mhi_cntrl)
 
 	dev_info(dev, "Requested to power ON\n");
 
-	if (mhi_cntrl->nr_irqs < 1)
-		return -EINVAL;
-
 	/* Supply default wake routines if not provided by controller driver */
 	if (!mhi_cntrl->wake_get || !mhi_cntrl->wake_put ||
 	    !mhi_cntrl->wake_toggle) {
-- 
2.25.1

[PATCH 22/29] bus: mhi: core: Separate system error and power down handling

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

Currently, there exist a set of if...else statements in the
mhi_pm_disable_transition() function which make handling system
error and disable transitions differently complex. To make that
cleaner and facilitate differences in behavior, separate these
two transitions for MHI host.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/pm.c | 159 ++++++++++++++++++++++++++++++++------
 1 file changed, 137 insertions(+), 22 deletions(-)

diff --git a/drivers/bus/mhi/core/pm.c b/drivers/bus/mhi/core/pm.c
index 1d04e401b67f..347ae7dd6cf8 100644
--- a/drivers/bus/mhi/core/pm.c
+++ b/drivers/bus/mhi/core/pm.c
@@ -444,7 +444,7 @@ static int mhi_pm_mission_mode_transition(struct mhi_controller *mhi_cntrl)
 	return ret;
 }
 
-/* Handle SYS_ERR and Shutdown transitions */
+/* Handle shutdown transitions */
 static void mhi_pm_disable_transition(struct mhi_controller *mhi_cntrl,
 				      enum mhi_pm_state transition_state)
 {
@@ -460,10 +460,6 @@ static void mhi_pm_disable_transition(struct mhi_controller *mhi_cntrl,
 		to_mhi_pm_state_str(mhi_cntrl->pm_state),
 		to_mhi_pm_state_str(transition_state));
 
-	/* We must notify MHI control driver so it can clean up first */
-	if (transition_state == MHI_PM_SYS_ERR_PROCESS)
-		mhi_cntrl->status_cb(mhi_cntrl, MHI_CB_SYS_ERROR);
-
 	mutex_lock(&mhi_cntrl->pm_mutex);
 	write_lock_irq(&mhi_cntrl->pm_lock);
 	prev_state = mhi_cntrl->pm_state;
@@ -502,11 +498,8 @@ static void mhi_pm_disable_transition(struct mhi_controller *mhi_cntrl,
 							    MHICTRL_RESET_SHIFT,
 							    &in_reset) ||
 					!in_reset, timeout);
-		if ((!ret || in_reset) && cur_state == MHI_PM_SYS_ERR_PROCESS) {
+		if (!ret || in_reset)
 			dev_err(dev, "Device failed to exit MHI Reset state\n");
-			mutex_unlock(&mhi_cntrl->pm_mutex);
-			return;
-		}
 
 		/*
 		 * Device will clear BHI_INTVEC as a part of RESET processing,
@@ -566,19 +559,142 @@ static void mhi_pm_disable_transition(struct mhi_controller *mhi_cntrl,
 		er_ctxt->wp = er_ctxt->rbase;
 	}
 
-	if (cur_state == MHI_PM_SYS_ERR_PROCESS) {
-		mhi_ready_state_transition(mhi_cntrl);
-	} else {
-		/* Move to disable state */
-		write_lock_irq(&mhi_cntrl->pm_lock);
-		cur_state = mhi_tryset_pm_state(mhi_cntrl, MHI_PM_DISABLE);
-		write_unlock_irq(&mhi_cntrl->pm_lock);
-		if (unlikely(cur_state != MHI_PM_DISABLE))
-			dev_err(dev, "Error moving from PM state: %s to: %s\n",
-				to_mhi_pm_state_str(cur_state),
-				to_mhi_pm_state_str(MHI_PM_DISABLE));
+	/* Move to disable state */
+	write_lock_irq(&mhi_cntrl->pm_lock);
+	cur_state = mhi_tryset_pm_state(mhi_cntrl, MHI_PM_DISABLE);
+	write_unlock_irq(&mhi_cntrl->pm_lock);
+	if (unlikely(cur_state != MHI_PM_DISABLE))
+		dev_err(dev, "Error moving from PM state: %s to: %s\n",
+			to_mhi_pm_state_str(cur_state),
+			to_mhi_pm_state_str(MHI_PM_DISABLE));
+
+	dev_dbg(dev, "Exiting with PM state: %s, MHI state: %s\n",
+		to_mhi_pm_state_str(mhi_cntrl->pm_state),
+		TO_MHI_STATE_STR(mhi_cntrl->dev_state));
+
+	mutex_unlock(&mhi_cntrl->pm_mutex);
+}
+
+/* Handle system error transitions */
+static void mhi_pm_sys_error_transition(struct mhi_controller *mhi_cntrl)
+{
+	enum mhi_pm_state cur_state, prev_state;
+	struct mhi_event *mhi_event;
+	struct mhi_cmd_ctxt *cmd_ctxt;
+	struct mhi_cmd *mhi_cmd;
+	struct mhi_event_ctxt *er_ctxt;
+	struct device *dev = &mhi_cntrl->mhi_dev->dev;
+	int ret, i;
+
+	dev_dbg(dev, "Transitioning from PM state: %s to: %s\n",
+		to_mhi_pm_state_str(mhi_cntrl->pm_state),
+		to_mhi_pm_state_str(MHI_PM_SYS_ERR_PROCESS));
+
+	/* We must notify MHI control driver so it can clean up first */
+	mhi_cntrl->status_cb(mhi_cntrl, MHI_CB_SYS_ERROR);
+
+	mutex_lock(&mhi_cntrl->pm_mutex);
+	write_lock_irq(&mhi_cntrl->pm_lock);
+	prev_state = mhi_cntrl->pm_state;
+	cur_state = mhi_tryset_pm_state(mhi_cntrl, MHI_PM_SYS_ERR_PROCESS);
+	write_unlock_irq(&mhi_cntrl->pm_lock);
+
+	if (cur_state != MHI_PM_SYS_ERR_PROCESS) {
+		dev_err(dev, "Failed to transition from PM state: %s to: %s\n",
+			to_mhi_pm_state_str(cur_state),
+			to_mhi_pm_state_str(MHI_PM_SYS_ERR_PROCESS));
+		goto exit_sys_error_transition;
+	}
+
+	mhi_cntrl->ee = MHI_EE_DISABLE_TRANSITION;
+	mhi_cntrl->dev_state = MHI_STATE_RESET;
+
+	/* Wake up threads waiting for state transition */
+	wake_up_all(&mhi_cntrl->state_event);
+
+	/* Trigger MHI RESET so that the device will not access host memory */
+	if (MHI_REG_ACCESS_VALID(prev_state)) {
+		u32 in_reset = -1;
+		unsigned long timeout = msecs_to_jiffies(mhi_cntrl->timeout_ms);
+
+		dev_dbg(dev, "Triggering MHI Reset in device\n");
+		mhi_set_mhi_state(mhi_cntrl, MHI_STATE_RESET);
+
+		/* Wait for the reset bit to be cleared by the device */
+		ret = wait_event_timeout(mhi_cntrl->state_event,
+					 mhi_read_reg_field(mhi_cntrl,
+							    mhi_cntrl->regs,
+							    MHICTRL,
+							    MHICTRL_RESET_MASK,
+							    MHICTRL_RESET_SHIFT,
+							    &in_reset) ||
+					!in_reset, timeout);
+		if (!ret || in_reset) {
+			dev_err(dev, "Device failed to exit MHI Reset state\n");
+			goto exit_sys_error_transition;
+		}
+
+		/*
+		 * Device will clear BHI_INTVEC as a part of RESET processing,
+		 * hence re-program it
+		 */
+		mhi_write_reg(mhi_cntrl, mhi_cntrl->bhi, BHI_INTVEC, 0);
+	}
+
+	dev_dbg(dev,
+		"Waiting for all pending event ring processing to complete\n");
+	mhi_event = mhi_cntrl->mhi_event;
+	for (i = 0; i < mhi_cntrl->total_ev_rings; i++, mhi_event++) {
+		if (mhi_event->offload_ev)
+			continue;
+		tasklet_kill(&mhi_event->task);
 	}
 
+	/* Release lock and wait for all pending threads to complete */
+	mutex_unlock(&mhi_cntrl->pm_mutex);
+	dev_dbg(dev, "Waiting for all pending threads to complete\n");
+	wake_up_all(&mhi_cntrl->state_event);
+
+	dev_dbg(dev, "Reset all active channels and remove MHI devices\n");
+	device_for_each_child(mhi_cntrl->cntrl_dev, NULL, mhi_destroy_device);
+
+	mutex_lock(&mhi_cntrl->pm_mutex);
+
+	WARN_ON(atomic_read(&mhi_cntrl->dev_wake));
+	WARN_ON(atomic_read(&mhi_cntrl->pending_pkts));
+
+	/* Reset the ev rings and cmd rings */
+	dev_dbg(dev, "Resetting EV CTXT and CMD CTXT\n");
+	mhi_cmd = mhi_cntrl->mhi_cmd;
+	cmd_ctxt = mhi_cntrl->mhi_ctxt->cmd_ctxt;
+	for (i = 0; i < NR_OF_CMD_RINGS; i++, mhi_cmd++, cmd_ctxt++) {
+		struct mhi_ring *ring = &mhi_cmd->ring;
+
+		ring->rp = ring->base;
+		ring->wp = ring->base;
+		cmd_ctxt->rp = cmd_ctxt->rbase;
+		cmd_ctxt->wp = cmd_ctxt->rbase;
+	}
+
+	mhi_event = mhi_cntrl->mhi_event;
+	er_ctxt = mhi_cntrl->mhi_ctxt->er_ctxt;
+	for (i = 0; i < mhi_cntrl->total_ev_rings; i++, er_ctxt++,
+	     mhi_event++) {
+		struct mhi_ring *ring = &mhi_event->ring;
+
+		/* Skip offload events */
+		if (mhi_event->offload_ev)
+			continue;
+
+		ring->rp = ring->base;
+		ring->wp = ring->base;
+		er_ctxt->rp = er_ctxt->rbase;
+		er_ctxt->wp = er_ctxt->rbase;
+	}
+
+	mhi_ready_state_transition(mhi_cntrl);
+
+exit_sys_error_transition:
 	dev_dbg(dev, "Exiting with PM state: %s, MHI state: %s\n",
 		to_mhi_pm_state_str(mhi_cntrl->pm_state),
 		TO_MHI_STATE_STR(mhi_cntrl->dev_state));
@@ -666,8 +782,7 @@ void mhi_pm_st_worker(struct work_struct *work)
 			mhi_ready_state_transition(mhi_cntrl);
 			break;
 		case DEV_ST_TRANSITION_SYS_ERR:
-			mhi_pm_disable_transition
-				(mhi_cntrl, MHI_PM_SYS_ERR_PROCESS);
+			mhi_pm_sys_error_transition(mhi_cntrl);
 			break;
 		case DEV_ST_TRANSITION_DISABLE:
 			mhi_pm_disable_transition
-- 
2.25.1

[PATCH 23/29] bus: mhi: core: Mark and maintain device states early on after power down

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

mhi_power_down() does not ensure that the PM state is moved to an
inaccessible state soon enough as the system can encounter
scheduling delays till mhi_pm_disable_transition() gets called.
Additionally, if an MHI controller decides that the device is now
inaccessible and issues a power down, the register inaccessible
state is not maintained by moving from MHI_PM_LD_ERR_FATAL_DETECT
to MHI_PM_SHUTDOWN_PROCESS. This can result in bus errors if a
client driver attempted to read registers when powering down.
Close these gaps and avoid any race conditions to prevent such
activity.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/pm.c | 77 +++++++++++++++++----------------------
 1 file changed, 33 insertions(+), 44 deletions(-)

diff --git a/drivers/bus/mhi/core/pm.c b/drivers/bus/mhi/core/pm.c
index 347ae7dd6cf8..ffbf6f539510 100644
--- a/drivers/bus/mhi/core/pm.c
+++ b/drivers/bus/mhi/core/pm.c
@@ -37,9 +37,10 @@
  *     M0 -> FW_DL_ERR
  *     M0 -> M3_ENTER -> M3 -> M3_EXIT --> M0
  * L1: SYS_ERR_DETECT -> SYS_ERR_PROCESS --> POR
- * L2: SHUTDOWN_PROCESS -> DISABLE
+ * L2: SHUTDOWN_PROCESS -> LD_ERR_FATAL_DETECT
+ *     SHUTDOWN_PROCESS -> DISABLE
  * L3: LD_ERR_FATAL_DETECT <--> LD_ERR_FATAL_DETECT
- *     LD_ERR_FATAL_DETECT -> SHUTDOWN_PROCESS
+ *     LD_ERR_FATAL_DETECT -> DISABLE
  */
 static struct mhi_pm_transitions const dev_state_transitions[] = {
 	/* L0 States */
@@ -72,7 +73,7 @@ static struct mhi_pm_transitions const dev_state_transitions[] = {
 	{
 		MHI_PM_M3,
 		MHI_PM_M3_EXIT | MHI_PM_SYS_ERR_DETECT |
-		MHI_PM_SHUTDOWN_PROCESS | MHI_PM_LD_ERR_FATAL_DETECT
+		MHI_PM_LD_ERR_FATAL_DETECT
 	},
 	{
 		MHI_PM_M3_EXIT,
@@ -103,7 +104,7 @@ static struct mhi_pm_transitions const dev_state_transitions[] = {
 	/* L3 States */
 	{
 		MHI_PM_LD_ERR_FATAL_DETECT,
-		MHI_PM_LD_ERR_FATAL_DETECT | MHI_PM_SHUTDOWN_PROCESS
+		MHI_PM_LD_ERR_FATAL_DETECT | MHI_PM_DISABLE
 	},
 };
 
@@ -445,10 +446,9 @@ static int mhi_pm_mission_mode_transition(struct mhi_controller *mhi_cntrl)
 }
 
 /* Handle shutdown transitions */
-static void mhi_pm_disable_transition(struct mhi_controller *mhi_cntrl,
-				      enum mhi_pm_state transition_state)
+static void mhi_pm_disable_transition(struct mhi_controller *mhi_cntrl)
 {
-	enum mhi_pm_state cur_state, prev_state;
+	enum mhi_pm_state cur_state;
 	struct mhi_event *mhi_event;
 	struct mhi_cmd_ctxt *cmd_ctxt;
 	struct mhi_cmd *mhi_cmd;
@@ -456,33 +456,13 @@ static void mhi_pm_disable_transition(struct mhi_controller *mhi_cntrl,
 	struct device *dev = &mhi_cntrl->mhi_dev->dev;
 	int ret, i;
 
-	dev_dbg(dev, "Transitioning from PM state: %s to: %s\n",
-		to_mhi_pm_state_str(mhi_cntrl->pm_state),
-		to_mhi_pm_state_str(transition_state));
+	dev_dbg(dev, "Processing disable transition with PM state: %s\n",
+		to_mhi_pm_state_str(mhi_cntrl->pm_state));
 
 	mutex_lock(&mhi_cntrl->pm_mutex);
-	write_lock_irq(&mhi_cntrl->pm_lock);
-	prev_state = mhi_cntrl->pm_state;
-	cur_state = mhi_tryset_pm_state(mhi_cntrl, transition_state);
-	if (cur_state == transition_state) {
-		mhi_cntrl->ee = MHI_EE_DISABLE_TRANSITION;
-		mhi_cntrl->dev_state = MHI_STATE_RESET;
-	}
-	write_unlock_irq(&mhi_cntrl->pm_lock);
-
-	/* Wake up threads waiting for state transition */
-	wake_up_all(&mhi_cntrl->state_event);
-
-	if (cur_state != transition_state) {
-		dev_err(dev, "Failed to transition to state: %s from: %s\n",
-			to_mhi_pm_state_str(transition_state),
-			to_mhi_pm_state_str(cur_state));
-		mutex_unlock(&mhi_cntrl->pm_mutex);
-		return;
-	}
 
 	/* Trigger MHI RESET so that the device will not access host memory */
-	if (MHI_REG_ACCESS_VALID(prev_state)) {
+	if (!MHI_PM_IN_FATAL_STATE(mhi_cntrl->pm_state)) {
 		u32 in_reset = -1;
 		unsigned long timeout = msecs_to_jiffies(mhi_cntrl->timeout_ms);
 
@@ -785,8 +765,7 @@ void mhi_pm_st_worker(struct work_struct *work)
 			mhi_pm_sys_error_transition(mhi_cntrl);
 			break;
 		case DEV_ST_TRANSITION_DISABLE:
-			mhi_pm_disable_transition
-				(mhi_cntrl, MHI_PM_SHUTDOWN_PROCESS);
+			mhi_pm_disable_transition(mhi_cntrl);
 			break;
 		default:
 			break;
@@ -1153,23 +1132,33 @@ EXPORT_SYMBOL_GPL(mhi_async_power_up);
 
 void mhi_power_down(struct mhi_controller *mhi_cntrl, bool graceful)
 {
-	enum mhi_pm_state cur_state;
+	enum mhi_pm_state cur_state, transition_state;
 	struct device *dev = &mhi_cntrl->mhi_dev->dev;
 
 	/* If it's not a graceful shutdown, force MHI to linkdown state */
-	if (!graceful) {
-		mutex_lock(&mhi_cntrl->pm_mutex);
-		write_lock_irq(&mhi_cntrl->pm_lock);
-		cur_state = mhi_tryset_pm_state(mhi_cntrl,
-						MHI_PM_LD_ERR_FATAL_DETECT);
-		write_unlock_irq(&mhi_cntrl->pm_lock);
-		mutex_unlock(&mhi_cntrl->pm_mutex);
-		if (cur_state != MHI_PM_LD_ERR_FATAL_DETECT)
-			dev_dbg(dev, "Failed to move to state: %s from: %s\n",
-				to_mhi_pm_state_str(MHI_PM_LD_ERR_FATAL_DETECT),
-				to_mhi_pm_state_str(mhi_cntrl->pm_state));
+	transition_state = (graceful) ? MHI_PM_SHUTDOWN_PROCESS :
+			   MHI_PM_LD_ERR_FATAL_DETECT;
+
+	mutex_lock(&mhi_cntrl->pm_mutex);
+	write_lock_irq(&mhi_cntrl->pm_lock);
+	cur_state = mhi_tryset_pm_state(mhi_cntrl, transition_state);
+	if (cur_state != transition_state) {
+		dev_err(dev, "Failed to move to state: %s from: %s\n",
+			to_mhi_pm_state_str(transition_state),
+			to_mhi_pm_state_str(mhi_cntrl->pm_state));
+		/* Force link down or error fatal detected state */
+		mhi_cntrl->pm_state = MHI_PM_LD_ERR_FATAL_DETECT;
 	}
 
+	/* mark device inactive to avoid any further host processing */
+	mhi_cntrl->ee = MHI_EE_DISABLE_TRANSITION;
+	mhi_cntrl->dev_state = MHI_STATE_RESET;
+
+	wake_up_all(&mhi_cntrl->state_event);
+
+	write_unlock_irq(&mhi_cntrl->pm_lock);
+	mutex_unlock(&mhi_cntrl->pm_mutex);
+
 	mhi_queue_state_transition(mhi_cntrl, DEV_ST_TRANSITION_DISABLE);
 
 	/* Wait for shutdown to complete */
-- 
2.25.1

[PATCH 24/29] bus: mhi: core: Remove MHI event ring IRQ handlers when powering down

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

While powering down, the device may or may not acknowledge an MHI
RESET issued by host for a graceful shutdown scenario and end up
sending an incoming data packet after tasklets have been killed.
If a rogue device sends this interrupt for a data transfer event
ring update, it can result in a tasklet getting scheduled while a
clean up is ongoing or has completed and cause access to freed
memory leading to a NULL pointer exception. Remove the interrupt
handlers for MHI event rings early on to avoid this scenario.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/pm.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/bus/mhi/core/pm.c b/drivers/bus/mhi/core/pm.c
index ffbf6f539510..a671f585ce35 100644
--- a/drivers/bus/mhi/core/pm.c
+++ b/drivers/bus/mhi/core/pm.c
@@ -494,6 +494,7 @@ static void mhi_pm_disable_transition(struct mhi_controller *mhi_cntrl)
 	for (i = 0; i < mhi_cntrl->total_ev_rings; i++, mhi_event++) {
 		if (mhi_event->offload_ev)
 			continue;
+		free_irq(mhi_cntrl->irq[mhi_event->irq], mhi_event);
 		tasklet_kill(&mhi_event->task);
 	}
 
@@ -1164,7 +1165,7 @@ void mhi_power_down(struct mhi_controller *mhi_cntrl, bool graceful)
 	/* Wait for shutdown to complete */
 	flush_work(&mhi_cntrl->st_worker);
 
-	mhi_deinit_free_irq(mhi_cntrl);
+	free_irq(mhi_cntrl->irq[0], mhi_cntrl);
 
 	if (!mhi_cntrl->pre_init) {
 		/* Free all allocated resources */
-- 
2.25.1

[PATCH 25/29] net: qrtr: Unprepare MHI channels during remove

[Manivannan Sadhasivam]

From: Bhaumik Bhatt <bbhatt@codeaurora.org>

Reset MHI device channels when driver remove is called due to
module unload or any crash scenario. This will make sure that
MHI channels no longer remain enabled for transfers since the
MHI stack does not take care of this anymore after the auto-start
channels feature was removed.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
Acked-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 net/qrtr/mhi.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/net/qrtr/mhi.c b/net/qrtr/mhi.c
index 7100f0bac4c6..2bf2b1943e61 100644
--- a/net/qrtr/mhi.c
+++ b/net/qrtr/mhi.c
@@ -104,6 +104,7 @@ static void qcom_mhi_qrtr_remove(struct mhi_device *mhi_dev)
 	struct qrtr_mhi_dev *qdev = dev_get_drvdata(&mhi_dev->dev);
 
 	qrtr_endpoint_unregister(&qdev->ep);
+	mhi_unprepare_from_transfer(mhi_dev);
 	dev_set_drvdata(&mhi_dev->dev, NULL);
 }
 
-- 
2.25.1

[PATCH 26/29] bus: mhi: core: Indexed MHI controller name

[Manivannan Sadhasivam]

From: Loic Poulain <loic.poulain@linaro.org>

Today the MHI controller name is simply cloned from the underlying
bus device (its parent), that gives the following device structure
for e.g. a MHI/PCI controller:
devices/pci0000:00/0000:00:01.2/0000:02:00.0/0000:02:00.0
devices/pci0000:00/0000:00:01.2/0000:02:00.0/0000:02:00.0/0000:02:00.0_IPCR
...

That's quite misleading/confusing and can cause device registering
issues because of duplicate dev name (e.g. if a PCI device register
two different MHI instances).

This patch changes MHI core to create indexed mhi controller names
(mhi0, mhi1...) in the same way as other busses (i2c0, usb0...).

The previous example becomes:
devices/pci0000:00/0000:00:01.2/0000:02:00.0/mhi0
devices/pci0000:00/0000:00:01.2/0000:02:00.0/mhi0/mhi0_IPCR
...
 v2: move index field at the end of mhi_controller struct (before bool)
     to avoid breaking well packed alignment.

Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Reviewed-by: Jeffrey Hugo <jhugo@codeaurora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/init.c | 18 ++++++++++++++++--
 drivers/bus/mhi/core/main.c |  2 +-
 include/linux/mhi.h         |  2 ++
 3 files changed, 19 insertions(+), 3 deletions(-)

diff --git a/drivers/bus/mhi/core/init.c b/drivers/bus/mhi/core/init.c
index 655d539c6808..1d6f7b6c1fcd 100644
--- a/drivers/bus/mhi/core/init.c
+++ b/drivers/bus/mhi/core/init.c
@@ -8,6 +8,7 @@
 #include <linux/device.h>
 #include <linux/dma-direction.h>
 #include <linux/dma-mapping.h>
+#include <linux/idr.h>
 #include <linux/interrupt.h>
 #include <linux/list.h>
 #include <linux/mhi.h>
@@ -18,6 +19,8 @@
 #include <linux/wait.h>
 #include "internal.h"
 
+static DEFINE_IDA(mhi_controller_ida);
+
 const char * const mhi_ee_str[MHI_EE_MAX] = {
 	[MHI_EE_PBL] = "PBL",
 	[MHI_EE_SBL] = "SBL",
@@ -940,6 +943,12 @@ int mhi_register_controller(struct mhi_controller *mhi_cntrl,
 	mhi_cntrl->minor_version = (soc_info & SOC_HW_VERSION_MINOR_VER_BMSK) >>
 					SOC_HW_VERSION_MINOR_VER_SHFT;
 
+	mhi_cntrl->index = ida_alloc(&mhi_controller_ida, GFP_KERNEL);
+	if (mhi_cntrl->index < 0) {
+		ret = mhi_cntrl->index;
+		goto error_ida_alloc;
+	}
+
 	/* Register controller with MHI bus */
 	mhi_dev = mhi_alloc_device(mhi_cntrl);
 	if (IS_ERR(mhi_dev)) {
@@ -950,8 +959,8 @@ int mhi_register_controller(struct mhi_controller *mhi_cntrl,
 
 	mhi_dev->dev_type = MHI_DEVICE_CONTROLLER;
 	mhi_dev->mhi_cntrl = mhi_cntrl;
-	dev_set_name(&mhi_dev->dev, "%s", dev_name(mhi_cntrl->cntrl_dev));
-	mhi_dev->name = dev_name(mhi_cntrl->cntrl_dev);
+	dev_set_name(&mhi_dev->dev, "mhi%d", mhi_cntrl->index);
+	mhi_dev->name = dev_name(&mhi_dev->dev);
 
 	/* Init wakeup source */
 	device_init_wakeup(&mhi_dev->dev, true);
@@ -970,6 +979,9 @@ int mhi_register_controller(struct mhi_controller *mhi_cntrl,
 	put_device(&mhi_dev->dev);
 
 error_alloc_dev:
+	ida_free(&mhi_controller_ida, mhi_cntrl->index);
+
+error_ida_alloc:
 	kfree(mhi_cntrl->mhi_cmd);
 
 error_alloc_cmd:
@@ -1004,6 +1016,8 @@ void mhi_unregister_controller(struct mhi_controller *mhi_cntrl)
 
 	device_del(&mhi_dev->dev);
 	put_device(&mhi_dev->dev);
+
+	ida_free(&mhi_controller_ida, mhi_cntrl->index);
 }
 EXPORT_SYMBOL_GPL(mhi_unregister_controller);
 
diff --git a/drivers/bus/mhi/core/main.c b/drivers/bus/mhi/core/main.c
index 4eb93d8bea1d..702c31b6aefa 100644
--- a/drivers/bus/mhi/core/main.c
+++ b/drivers/bus/mhi/core/main.c
@@ -331,7 +331,7 @@ void mhi_create_devices(struct mhi_controller *mhi_cntrl)
 		/* Channel name is same for both UL and DL */
 		mhi_dev->name = mhi_chan->name;
 		dev_set_name(&mhi_dev->dev, "%s_%s",
-			     dev_name(mhi_cntrl->cntrl_dev),
+			     dev_name(&mhi_cntrl->mhi_dev->dev),
 			     mhi_dev->name);
 
 		/* Init wakeup source if available */
diff --git a/include/linux/mhi.h b/include/linux/mhi.h
index d31efcf02ae7..aa9757e71f1f 100644
--- a/include/linux/mhi.h
+++ b/include/linux/mhi.h
@@ -348,6 +348,7 @@ struct mhi_controller_config {
  * @read_reg: Read a MHI register via the physical link (required)
  * @write_reg: Write a MHI register via the physical link (required)
  * @buffer_len: Bounce buffer length
+ * @index: Index of the MHI controller instance
  * @bounce_buf: Use of bounce buffer
  * @fbc_download: MHI host needs to do complete image transfer (optional)
  * @pre_init: MHI host needs to do pre-initialization before power up
@@ -438,6 +439,7 @@ struct mhi_controller {
 			  u32 val);
 
 	size_t buffer_len;
+	int index;
 	bool bounce_buf;
 	bool fbc_download;
 	bool pre_init;
-- 
2.25.1

[PATCH 27/29] bus: mhi: core: Fix device hierarchy

[Manivannan Sadhasivam]

From: Loic Poulain <loic.poulain@linaro.org>

This patch fixes the hierarchical structure of MHI devices. Indeed,
MHI client devices are directly 'enumerated' from the mhi controller
and therefore must be direct descendants/children of their mhi
controller device, in accordance with the Linux Device Model.

Today both MHI clients and controller devices are at the same level,
this patch ensures that MHI controller is parent of its client devices.

The hierarchy is especially important for power management (safe
suspend/resume order). It is also useful for userspace to determine
relationship between MHI client devices and controllers.

Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Reviewed-by: Hemant Kumar <hemantk@codeaurora.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/debugfs.c |  4 +++-
 drivers/bus/mhi/core/init.c    | 10 +++++++++-
 drivers/bus/mhi/core/pm.c      |  4 ++--
 3 files changed, 14 insertions(+), 4 deletions(-)

diff --git a/drivers/bus/mhi/core/debugfs.c b/drivers/bus/mhi/core/debugfs.c
index 3a48801e01f4..7d43138ce66d 100644
--- a/drivers/bus/mhi/core/debugfs.c
+++ b/drivers/bus/mhi/core/debugfs.c
@@ -159,7 +159,9 @@ static int mhi_debugfs_devices_show(struct seq_file *m, void *d)
 		return -ENODEV;
 	}
 
-	device_for_each_child(mhi_cntrl->cntrl_dev, m, mhi_device_info_show);
+	/* Show controller and client(s) info */
+	mhi_device_info_show(&mhi_cntrl->mhi_dev->dev, m);
+	device_for_each_child(&mhi_cntrl->mhi_dev->dev, m, mhi_device_info_show);
 
 	return 0;
 }
diff --git a/drivers/bus/mhi/core/init.c b/drivers/bus/mhi/core/init.c
index 1d6f7b6c1fcd..96cde9c0034c 100644
--- a/drivers/bus/mhi/core/init.c
+++ b/drivers/bus/mhi/core/init.c
@@ -1144,7 +1144,15 @@ struct mhi_device *mhi_alloc_device(struct mhi_controller *mhi_cntrl)
 	device_initialize(dev);
 	dev->bus = &mhi_bus_type;
 	dev->release = mhi_release_device;
-	dev->parent = mhi_cntrl->cntrl_dev;
+
+	if (mhi_cntrl->mhi_dev) {
+		/* for MHI client devices, parent is the MHI controller device */
+		dev->parent = &mhi_cntrl->mhi_dev->dev;
+	} else {
+		/* for MHI controller device, parent is the bus device (e.g. pci device) */
+		dev->parent = mhi_cntrl->cntrl_dev;
+	}
+
 	mhi_dev->mhi_cntrl = mhi_cntrl;
 	mhi_dev->dev_wake = 0;
 
diff --git a/drivers/bus/mhi/core/pm.c b/drivers/bus/mhi/core/pm.c
index a671f585ce35..681960c72d2a 100644
--- a/drivers/bus/mhi/core/pm.c
+++ b/drivers/bus/mhi/core/pm.c
@@ -504,7 +504,7 @@ static void mhi_pm_disable_transition(struct mhi_controller *mhi_cntrl)
 	wake_up_all(&mhi_cntrl->state_event);
 
 	dev_dbg(dev, "Reset all active channels and remove MHI devices\n");
-	device_for_each_child(mhi_cntrl->cntrl_dev, NULL, mhi_destroy_device);
+	device_for_each_child(&mhi_cntrl->mhi_dev->dev, NULL, mhi_destroy_device);
 
 	mutex_lock(&mhi_cntrl->pm_mutex);
 
@@ -637,7 +637,7 @@ static void mhi_pm_sys_error_transition(struct mhi_controller *mhi_cntrl)
 	wake_up_all(&mhi_cntrl->state_event);
 
 	dev_dbg(dev, "Reset all active channels and remove MHI devices\n");
-	device_for_each_child(mhi_cntrl->cntrl_dev, NULL, mhi_destroy_device);
+	device_for_each_child(&mhi_cntrl->mhi_dev->dev, NULL, mhi_destroy_device);
 
 	mutex_lock(&mhi_cntrl->pm_mutex);
 
-- 
2.25.1

[PATCH 28/29] bus: mhi: core: Fix error handling in mhi_register_controller()

[Manivannan Sadhasivam]

From: Dan Carpenter <dan.carpenter@oracle.com>

There are a few problems with the error handling in this function.  They
mostly center around the alloc_ordered_workqueue() allocation.
1) If that allocation fails or if the kcalloc() prior to it fails then
it leads to a NULL dereference when we call
destroy_workqueue(mhi_cntrl->hiprio_wq).
2) The error code is not set.
3) The "mhi_cntrl->mhi_cmd" allocation is not freed.

The error handling was slightly confusing and I re-ordered it to be in
the exact mirror/reverse order of how things were allocated.  I changed
the label names to say what the goto does instead of describing where
the goto comes from.

Fixes: 8f7039787687 ("bus: mhi: core: Move to using high priority workqueue")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Loic Poulain <loic.poulain@linaro.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/core/init.c | 29 ++++++++++++++---------------
 1 file changed, 14 insertions(+), 15 deletions(-)

diff --git a/drivers/bus/mhi/core/init.c b/drivers/bus/mhi/core/init.c
index 96cde9c0034c..f0697f433c2f 100644
--- a/drivers/bus/mhi/core/init.c
+++ b/drivers/bus/mhi/core/init.c
@@ -871,7 +871,7 @@ int mhi_register_controller(struct mhi_controller *mhi_cntrl,
 				     sizeof(*mhi_cntrl->mhi_cmd), GFP_KERNEL);
 	if (!mhi_cntrl->mhi_cmd) {
 		ret = -ENOMEM;
-		goto error_alloc_cmd;
+		goto err_free_event;
 	}
 
 	INIT_LIST_HEAD(&mhi_cntrl->transition_list);
@@ -886,7 +886,8 @@ int mhi_register_controller(struct mhi_controller *mhi_cntrl,
 				("mhi_hiprio_wq", WQ_MEM_RECLAIM | WQ_HIGHPRI);
 	if (!mhi_cntrl->hiprio_wq) {
 		dev_err(mhi_cntrl->cntrl_dev, "Failed to allocate workqueue\n");
-		goto error_alloc_cmd;
+		ret = -ENOMEM;
+		goto err_free_cmd;
 	}
 
 	mhi_cmd = mhi_cntrl->mhi_cmd;
@@ -932,7 +933,7 @@ int mhi_register_controller(struct mhi_controller *mhi_cntrl,
 	ret = mhi_read_reg(mhi_cntrl, mhi_cntrl->regs,
 			   SOC_HW_VERSION_OFFS, &soc_info);
 	if (ret)
-		goto error_alloc_dev;
+		goto err_destroy_wq;
 
 	mhi_cntrl->family_number = (soc_info & SOC_HW_VERSION_FAM_NUM_BMSK) >>
 					SOC_HW_VERSION_FAM_NUM_SHFT;
@@ -946,7 +947,7 @@ int mhi_register_controller(struct mhi_controller *mhi_cntrl,
 	mhi_cntrl->index = ida_alloc(&mhi_controller_ida, GFP_KERNEL);
 	if (mhi_cntrl->index < 0) {
 		ret = mhi_cntrl->index;
-		goto error_ida_alloc;
+		goto err_destroy_wq;
 	}
 
 	/* Register controller with MHI bus */
@@ -954,7 +955,7 @@ int mhi_register_controller(struct mhi_controller *mhi_cntrl,
 	if (IS_ERR(mhi_dev)) {
 		dev_err(mhi_cntrl->cntrl_dev, "Failed to allocate MHI device\n");
 		ret = PTR_ERR(mhi_dev);
-		goto error_alloc_dev;
+		goto err_ida_free;
 	}
 
 	mhi_dev->dev_type = MHI_DEVICE_CONTROLLER;
@@ -967,7 +968,7 @@ int mhi_register_controller(struct mhi_controller *mhi_cntrl,
 
 	ret = device_add(&mhi_dev->dev);
 	if (ret)
-		goto error_add_dev;
+		goto err_release_dev;
 
 	mhi_cntrl->mhi_dev = mhi_dev;
 
@@ -975,19 +976,17 @@ int mhi_register_controller(struct mhi_controller *mhi_cntrl,
 
 	return 0;
 
-error_add_dev:
+err_release_dev:
 	put_device(&mhi_dev->dev);
-
-error_alloc_dev:
+err_ida_free:
 	ida_free(&mhi_controller_ida, mhi_cntrl->index);
-
-error_ida_alloc:
+err_destroy_wq:
+	destroy_workqueue(mhi_cntrl->hiprio_wq);
+err_free_cmd:
 	kfree(mhi_cntrl->mhi_cmd);
-
-error_alloc_cmd:
-	vfree(mhi_cntrl->mhi_chan);
+err_free_event:
 	kfree(mhi_cntrl->mhi_event);
-	destroy_workqueue(mhi_cntrl->hiprio_wq);
+	vfree(mhi_cntrl->mhi_chan);
 
 	return ret;
 }
-- 
2.25.1

[PATCH 29/29] mhi: pci_generic: Fix implicit conversion warning

[Manivannan Sadhasivam]

From: Loic Poulain <loic.poulain@linaro.org>

Fix the following warning with explicit cast:

warning: implicit conversion from 'unsigned long long' to
'dma_addr_t' (aka 'unsigned int')
mhi_cntrl->iova_stop = DMA_BIT_MASK(info->dma_data_width);

Fixes: 855a70c12021 ("bus: mhi: Add MHI PCI support for WWAN modems")
Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Reported-by: kernel test robot <lkp@intel.com>
Reviewed-by: Nathan Chancellor <natechancellor@gmail.com>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/bus/mhi/pci_generic.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/bus/mhi/pci_generic.c b/drivers/bus/mhi/pci_generic.c
index e3df838c3c80..f5bee76ea061 100644
--- a/drivers/bus/mhi/pci_generic.c
+++ b/drivers/bus/mhi/pci_generic.c
@@ -273,7 +273,7 @@ static int mhi_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	mhi_cntrl_config = info->config;
 	mhi_cntrl->cntrl_dev = &pdev->dev;
 	mhi_cntrl->iova_start = 0;
-	mhi_cntrl->iova_stop = DMA_BIT_MASK(info->dma_data_width);
+	mhi_cntrl->iova_stop = (dma_addr_t)DMA_BIT_MASK(info->dma_data_width);
 	mhi_cntrl->fw_image = info->fw;
 	mhi_cntrl->edl_image = info->edl;
 
-- 
2.25.1

Re: [PATCH 01/29] bus: mhi: Remove auto-start option

[Kalle Valo]

Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> writes:

> From: Loic Poulain <loic.poulain@linaro.org>
>
> There is really no point having an auto-start for channels.
> This is confusing for the device drivers, some have to enable the
> channels, others don't have... and waste resources (e.g. pre allocated
> buffers) that may never be used.
>
> This is really up to the MHI device(channel) driver to manage the state
> of its channels.
>
> While at it, let's also remove the auto-start option from ath11k mhi
> controller.
>
> Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
> Acked-by: Kalle Valo <kvalo@codeaurora.org>
> Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
> [mani: clubbed ath11k change]
> Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>

Hmm, didn't we apply this already? At least I pulled the immutable
branch to my tree:

https://git.kernel.org/pub/scm/linux/kernel/git/kvalo/ath.git/commit/?h=ath-next&id=526740b495059ebbc0c3c086dceca1263820fa4f

So if there's a new version of this patch (and a new commit id) the
immutable branch I pulled is not immutable anymore.

-- 
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

Re: [PATCH 01/29] bus: mhi: Remove auto-start option

[Manivannan Sadhasivam]

On Wed, Dec 02, 2020 at 06:00:05PM +0200, Kalle Valo wrote:
> Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> writes:
> 
> > From: Loic Poulain <loic.poulain@linaro.org>
> >
> > There is really no point having an auto-start for channels.
> > This is confusing for the device drivers, some have to enable the
> > channels, others don't have... and waste resources (e.g. pre allocated
> > buffers) that may never be used.
> >
> > This is really up to the MHI device(channel) driver to manage the state
> > of its channels.
> >
> > While at it, let's also remove the auto-start option from ath11k mhi
> > controller.
> >
> > Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
> > Acked-by: Kalle Valo <kvalo@codeaurora.org>
> > Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
> > [mani: clubbed ath11k change]
> > Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
> 
> Hmm, didn't we apply this already? At least I pulled the immutable
> branch to my tree:
> 
> https://git.kernel.org/pub/scm/linux/kernel/git/kvalo/ath.git/commit/?h=ath-next&id=526740b495059ebbc0c3c086dceca1263820fa4f
> 
> So if there's a new version of this patch (and a new commit id) the
> immutable branch I pulled is not immutable anymore.

This is not a new version of the patch. The commit SHA of this patch in
immutable branch is ed5298c7d500abaf34ed7783969e953a1f028e5b and that is same
in mhi-next as well.

Now I'm funneling all patches in mhi-next to mainline through Greg, so this
patch is part of the (MHI changes for v5.11) patch series. Perhaps you might be
dealing with pull requests to Dave/Jakub but since the MHI patch flow is usually
fairly low, Greg is happy with patch series.

But since we are dealing with immutable branch I should send the pull request
to Greg now.

Thanks,
Mani

> 
> -- 
> https://patchwork.kernel.org/project/linux-wireless/list/
> 
> https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

Re: [PATCH 00/29] MHI changes for v5.11

[Manivannan Sadhasivam]

Hi Greg,

On Wed, Dec 02, 2020 at 03:11:30PM +0530, Manivannan Sadhasivam wrote:
> Hi Greg,
> 
> Here is the MHI patch series for v5.11. It was a busy cycle for us and so far we
> have accumulated almost 29 patches. Most of the patches are cleanups and fixes
> but there are some noticeable changes too:
> 
> 1. Loic finally removed the auto-start option from the channel parameters of the
> MHI controller. It is the duty of the client drivers like qrtr to start/stop the
> channels when required, so we decided to remove this option. As a side effect,
> we changed the qrtr driver to start the channels during its probe and removed
> the auto-start option from ath11k controller.
> 
> **NOTE** Since these changes spawns both MHI and networking trees, the patches
> are maintained in an immutable branch [1] and pulled into both mhi-next and
> ath11k-next branches. The networking patches got acks from ath11k and networking
> maintainers as well.
> 

Sigh... this is not going to work for patches in email form due to the immutable
branch. I will send a pull request this time.

Please ignore this series.

Thanks,
Mani

> 2. Loic added a generic MHI pci controller driver. This driver will be used by
> the PCI based Qualcomm modems like SDX55 and exposes channels such as QMI,
> IP_HW0, IPCR etc...
> 
> 3. Loic fixed the MHI device hierarchy by maintaining the correct parent child
> relationships. Earlier all MHI devices lived in the same level under the parent
> device like PCIe. But now, the MHI devices belonging to channels will become the
> children of controller MHI device.
> 
> 4. Finally Loic also improved the MHI device naming by using indexed names such
> as mhi0, mhi1, etc... This will break the userspace applications depending on
> the old naming convention but since the only one user so far is Jeff Hugo's AI
> accelerator apps, we decided to make this change now itself with his agreement.
> 
> 5. Bhaumik fixed the qrtr driver by stopping the channels during remove. This
> patch also got ack from networking maintainer and we decided to take it through
> MHI tree (via immutable branch) since we already had a qrtr change.
> 
> Please consider applying!
> 
> Thanks,
> Mani
> 
> [1] https://git.kernel.org/pub/scm/linux/kernel/git/mani/mhi.git/log/?h=mhi-ath11k-immutable
> 
> Bhaumik Bhatt (19):
>   bus: mhi: core: Remove unnecessary counter from mhi_firmware_copy()
>   bus: mhi: core: Add missing EXPORT_SYMBOL for mhi_get_mhi_state()
>   bus: mhi: core: Expose mhi_get_exec_env() API for controllers
>   bus: mhi: core: Remove unused mhi_fw_load_worker() declaration
>   bus: mhi: core: Rename RDDM download function to use proper words
>   bus: mhi: core: Skip RDDM download for unknown execution environment
>   bus: mhi: core: Use appropriate names for firmware load functions
>   bus: mhi: core: Move to using high priority workqueue
>   bus: mhi: core: Skip device wake in error or shutdown states
>   bus: mhi: core: Move to SYS_ERROR regardless of RDDM capability
>   bus: mhi: core: Prevent sending multiple RDDM entry callbacks
>   bus: mhi: core: Move to an error state on any firmware load failure
>   bus: mhi: core: Use appropriate label in firmware load handler API
>   bus: mhi: core: Move to an error state on mission mode failure
>   bus: mhi: core: Check for IRQ availability during registration
>   bus: mhi: core: Separate system error and power down handling
>   bus: mhi: core: Mark and maintain device states early on after power
>     down
>   bus: mhi: core: Remove MHI event ring IRQ handlers when powering down
>   net: qrtr: Unprepare MHI channels during remove
> 
> Carl Yin (1):
>   bus: mhi: core: Fix null pointer access when parsing MHI configuration
> 
> Dan Carpenter (1):
>   bus: mhi: core: Fix error handling in mhi_register_controller()
> 
> Jeffrey Hugo (1):
>   bus: mhi: core: fix potential operator-precedence with BHI macros
> 
> Loic Poulain (7):
>   bus: mhi: Remove auto-start option
>   net: qrtr: Start MHI channels during init
>   bus: mhi: Add MHI PCI support for WWAN modems
>   bus: mhi: Fix channel close issue on driver remove
>   bus: mhi: core: Indexed MHI controller name
>   bus: mhi: core: Fix device hierarchy
>   mhi: pci_generic: Fix implicit conversion warning
> 
> -- 
> 2.25.1
> 

Re: [PATCH 01/29] bus: mhi: Remove auto-start option

[Kalle Valo]

Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> writes:

> On Wed, Dec 02, 2020 at 06:00:05PM +0200, Kalle Valo wrote:
>> Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> writes:
>> 
>> > From: Loic Poulain <loic.poulain@linaro.org>
>> >
>> > There is really no point having an auto-start for channels.
>> > This is confusing for the device drivers, some have to enable the
>> > channels, others don't have... and waste resources (e.g. pre allocated
>> > buffers) that may never be used.
>> >
>> > This is really up to the MHI device(channel) driver to manage the state
>> > of its channels.
>> >
>> > While at it, let's also remove the auto-start option from ath11k mhi
>> > controller.
>> >
>> > Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
>> > Acked-by: Kalle Valo <kvalo@codeaurora.org>
>> > Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
>> > [mani: clubbed ath11k change]
>> > Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
>> 
>> Hmm, didn't we apply this already? At least I pulled the immutable
>> branch to my tree:
>> 
>> https://git.kernel.org/pub/scm/linux/kernel/git/kvalo/ath.git/commit/?h=ath-next&id=526740b495059ebbc0c3c086dceca1263820fa4f
>> 
>> So if there's a new version of this patch (and a new commit id) the
>> immutable branch I pulled is not immutable anymore.
>
> This is not a new version of the patch. The commit SHA of this patch in
> immutable branch is ed5298c7d500abaf34ed7783969e953a1f028e5b and that is same
> in mhi-next as well.
>
> Now I'm funneling all patches in mhi-next to mainline through Greg, so this
> patch is part of the (MHI changes for v5.11) patch series. Perhaps you might be
> dealing with pull requests to Dave/Jakub but since the MHI patch flow is usually
> fairly low, Greg is happy with patch series.
>
> But since we are dealing with immutable branch I should send the pull request
> to Greg now.

Yes, I very much prefer you send a pull request so that we can avoid
conflicts between ath11k and mhi. If Greg would apply this patch as is,
we would have two versions of the same commit (with different commit
ids) and git would get confused.

-- 
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

Re: [PATCH 01/29] bus: mhi: Remove auto-start option

[Greg KH]

On Wed, Dec 02, 2020 at 06:57:15PM +0200, Kalle Valo wrote:
> Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> writes:
> 
> > On Wed, Dec 02, 2020 at 06:00:05PM +0200, Kalle Valo wrote:
> >> Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> writes:
> >> 
> >> > From: Loic Poulain <loic.poulain@linaro.org>
> >> >
> >> > There is really no point having an auto-start for channels.
> >> > This is confusing for the device drivers, some have to enable the
> >> > channels, others don't have... and waste resources (e.g. pre allocated
> >> > buffers) that may never be used.
> >> >
> >> > This is really up to the MHI device(channel) driver to manage the state
> >> > of its channels.
> >> >
> >> > While at it, let's also remove the auto-start option from ath11k mhi
> >> > controller.
> >> >
> >> > Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
> >> > Acked-by: Kalle Valo <kvalo@codeaurora.org>
> >> > Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
> >> > [mani: clubbed ath11k change]
> >> > Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
> >> 
> >> Hmm, didn't we apply this already? At least I pulled the immutable
> >> branch to my tree:
> >> 
> >> https://git.kernel.org/pub/scm/linux/kernel/git/kvalo/ath.git/commit/?h=ath-next&id=526740b495059ebbc0c3c086dceca1263820fa4f
> >> 
> >> So if there's a new version of this patch (and a new commit id) the
> >> immutable branch I pulled is not immutable anymore.
> >
> > This is not a new version of the patch. The commit SHA of this patch in
> > immutable branch is ed5298c7d500abaf34ed7783969e953a1f028e5b and that is same
> > in mhi-next as well.
> >
> > Now I'm funneling all patches in mhi-next to mainline through Greg, so this
> > patch is part of the (MHI changes for v5.11) patch series. Perhaps you might be
> > dealing with pull requests to Dave/Jakub but since the MHI patch flow is usually
> > fairly low, Greg is happy with patch series.
> >
> > But since we are dealing with immutable branch I should send the pull request
> > to Greg now.
> 
> Yes, I very much prefer you send a pull request so that we can avoid
> conflicts between ath11k and mhi. If Greg would apply this patch as is,
> we would have two versions of the same commit (with different commit
> ids) and git would get confused.

git wouldn't get confused, we do merges like this all the time.

It would confuse developers, I get confused by this type of thing all
the time :)

I'll just do the pull instead to keep my sanity, git is fine...

thanks,

greg k-h