Skip to contentSkip to conversation

TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

PATCH In review Mathias Leblanc 4 participants 30 messages Nov 19, 2012 – Jan 22, 2013 1 message not in the archive

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

* STMicroelectronics version 1.2.0, Copyright (C) 2010
* STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
* This is free software, and you are welcome to redistribute it
* under certain conditions.

This is the driver for TPM chip from ST Microelectronics.

If you have a TPM security chip from STMicroelectronics working with an I2C, in menuconfig or .config choose the tpm driver on device --> tpm and activate the protocol of your choice before compiling the kernel.
The driver will be accessible from within Linux.

Tested on linux x86/x64, beagleboard REV B & XM REV C and CHROMIUM OS on kernel 3.x.x

---
 drivers/char/tpm/tpm_stm_st33_i2c.c  |  903 ++++++++++++++++++++++++++++++++++
 drivers/char/tpm/tpm_stm_st33_i2c.h  |   81 +++
 include/linux/i2c/tpm_stm_st33_i2c.h |   47 ++
 3 files changed, 1031 insertions(+)
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.c
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.h
 create mode 100644 include/linux/i2c/tpm_stm_st33_i2c.h
drivers/char/tpm/tpm_stm_st33_i2c.c new file +903 -0
new file mode 100644
index 0000000..3ec6ae9
@@ -0,0 +1,903 @@
+/*
+ * STMicroelectronics TPM I2C Linux driver for TPM ST33ZP24
+ * Copyright (C) 2009, 2010  STMicroelectronics
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * STMicroelectronics version 1.2.0, Copyright (C) 2010
+ * STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
+ * This is free software, and you are welcome to redistribute it
+ * under certain conditions.
+ *
+ * @Author: Christophe RICARD [email protected]
+ *
+ * @File: tpm_stm_st33_i2c.c
+ *
+ * @Synopsis:
+ *	09/15/2010:	First shot driver tpm_tis driver for
+			 lpc is used as model.
+ */
+
+#include <linux/i2c/tpm_stm_st33_i2c.h>
+
+#include "tpm_stm_st33_i2c.h"
+
+enum stm33zp24_access {
+	TPM_ACCESS_VALID = 0x80,
+	TPM_ACCESS_ACTIVE_LOCALITY = 0x20,
+	TPM_ACCESS_REQUEST_PENDING = 0x04,
+	TPM_ACCESS_REQUEST_USE = 0x02,
+};
+
+enum stm33zp24_status {
+	TPM_STS_VALID = 0x80,
+	TPM_STS_COMMAND_READY = 0x40,
+	TPM_STS_GO = 0x20,
+	TPM_STS_DATA_AVAIL = 0x10,
+	TPM_STS_DATA_EXPECT = 0x08,
+};
+
+enum stm33zp24_int_flags {
+	TPM_GLOBAL_INT_ENABLE = 0x80,
+	TPM_INTF_CMD_READY_INT = 0x080,
+	TPM_INTF_FIFO_AVALAIBLE_INT = 0x040,
+	TPM_INTF_WAKE_UP_READY_INT = 0x020,
+	TPM_INTF_LOCALITY_CHANGE_INT = 0x004,
+	TPM_INTF_STS_VALID_INT = 0x002,
+	TPM_INTF_DATA_AVAIL_INT = 0x001,
+};
+
+enum tis_defaults {
+	TIS_SHORT_TIMEOUT = 750,
+	TIS_LONG_TIMEOUT = 2000,
+};
+
+/*
+ * write8_reg
+ * Send byte to the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: tpm_register, the tpm tis register where the data should be written
+ * @param: tpm_data, the tpm_data to write inside the tpm_register
+ * @param: tpm_size, The length of the data
+ * @return: Returns negative errno, or else the number of bytes written.
+ */
+static int write8_reg(struct i2c_client *client, u8 tpm_register,
+		      u8 *tpm_data, u16 tpm_size)
+{
+	u8 data;
+	int value = 0;
+	struct st33zp24_platform_data *pin_infos;
+
+	pin_infos = client->dev.platform_data;
+
+	data = tpm_register;
+	memcpy(pin_infos->tpm_i2c_buffer[0], &data, sizeof(data));
+	memcpy(pin_infos->tpm_i2c_buffer[0] + 1, tpm_data, tpm_size);
+	value = i2c_master_send(client, pin_infos->tpm_i2c_buffer[0],
+				tpm_size + 1);
+	return value;
+} /* write8_reg() */
+
+/*
+ * read8_reg
+ * Recv byte from the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: tpm_register, the tpm tis register where the data should be read
+ * @param: tpm_data, the TPM response
+ * @param: tpm_size, tpm TPM response size to read.
+ * @return: number of byte read successfully: should be one if success.
+ */
+static int read8_reg(struct i2c_client *client, u8 tpm_register,
+		    u8 *tpm_data, int tpm_size)
+{
+	u8 status = 0;
+	u8 data;
+	struct st33zp24_platform_data *pin_infos;
+
+	pin_infos = client->dev.platform_data;
+
+	data = TPM_DUMMY_BYTE;
+	status = write8_reg(client, tpm_register, &data, 1);
+	if (status == 2)
+		status = i2c_master_recv(client, tpm_data, tpm_size);
+	return status;
+} /* read8_reg() */
+
+/*
+ * I2C_WRITE_DATA
+ * Send byte to the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: client, the chip description
+ * @param: tpm_register, the tpm tis register where the data should be written
+ * @param: tpm_data, the tpm_data to write inside the tpm_register
+ * @param: tpm_size, The length of the data
+ * @return: number of byte written successfully: should be one if success.
+ */
+#define I2C_WRITE_DATA(client, tpm_register, tpm_data, tpm_size) \
+	(write8_reg(client, tpm_register | \
+	TPM_WRITE_DIRECTION, tpm_data, tpm_size))
+
+/*
+ * I2C_READ_DATA
+ * Recv byte from the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: tpm, the chip description
+ * @param: tpm_register, the tpm tis register where the data should be read
+ * @param: tpm_data, the TPM response
+ * @param: tpm_size, tpm TPM response size to read.
+ * @return: number of byte read successfully: should be one if success.
+ */
+#define I2C_READ_DATA(client, tpm_register, tpm_data, tpm_size) \
+	(read8_reg(client, tpm_register, tpm_data, tpm_size))
+
+/*
+ * clear_interruption
+ * clear the TPM interrupt register.
+ * @param: tpm, the chip description
+ */
+static void clear_interruption(struct i2c_client *client)
+{
+	u8 interrupt;
+	I2C_READ_DATA(client, TPM_INT_STATUS, &interrupt, 1);
+	I2C_WRITE_DATA(client, TPM_INT_STATUS, &interrupt, 1);
+	I2C_READ_DATA(client, TPM_INT_STATUS, &interrupt, 1);
+} /* clear_interruption() */
+
+/*
+ * _wait_for_interrupt_serirq_timeout
+ * @param: tpm, the chip description
+ * @param: timeout, the timeout of the interrupt
+ * @return: the status of the interruption.
+ */
+static long _wait_for_interrupt_serirq_timeout(struct tpm_chip *chip,
+						unsigned long timeout)
+{
+	long status;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	status = wait_for_completion_interruptible_timeout(
+					&pin_infos->irq_detection,
+						timeout);
+	if (status > 0)
+		enable_irq(gpio_to_irq(pin_infos->io_serirq));
+	gpio_direction_input(pin_infos->io_serirq);
+
+	return status;
+} /* wait_for_interrupt_serirq_timeout() */
+
+int wait_for_serirq_timeout(struct tpm_chip *chip, bool condition,
+				 unsigned long timeout)
+{
+	int status = 2;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	status = _wait_for_interrupt_serirq_timeout(chip, timeout);
+	if (!status) {
+		status = -EBUSY;
+	} else{
+		clear_interruption(client);
+		if (condition)
+			status = 1;
+	}
+	return status;
+}
+
+/*
+ * tpm_stm_i2c_cancel, cancel is not implemented.
+ * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h
+ */
+static void tpm_stm_i2c_cancel(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	data = TPM_STS_COMMAND_READY;
+	I2C_WRITE_DATA(client, TPM_STS, &data, 1);
+	if (chip->vendor.irq)
+		wait_for_serirq_timeout(chip, 1, chip->vendor.timeout_a);
+}	/* tpm_stm_i2c_cancel() */
+
+/*
+ * tpm_stm_spi_status return the TPM_STS register
+ * @param: chip, the tpm chip description
+ * @return: the TPM_STS register value.
+ */
+static u8 tpm_stm_i2c_status(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	I2C_READ_DATA(client, TPM_STS, &data, 1);
+	return data;
+}				/* tpm_stm_i2c_status() */
+
+
+/*
+ * check_locality if the locality is active
+ * @param: chip, the tpm chip description
+ * @return: the active locality or -EACCESS.
+ */
+static int check_locality(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+	u8 status;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	status = I2C_READ_DATA(client, TPM_ACCESS, &data, 1);
+	if (status && (data &
+		(TPM_ACCESS_ACTIVE_LOCALITY | TPM_ACCESS_VALID)) ==
+		(TPM_ACCESS_ACTIVE_LOCALITY | TPM_ACCESS_VALID))
+		return chip->vendor.locality;
+
+	return -EACCES;
+
+} /* check_locality() */
+
+/*
+ * request_locality request the TPM locality
+ * @param: chip, the chip description
+ * @return: the active locality or EACCESS.
+ */
+static int request_locality(struct tpm_chip *chip)
+{
+	unsigned long stop;
+	long rc;
+	struct i2c_client *client;
+	u8 data;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	if (check_locality(chip) == chip->vendor.locality)
+		return chip->vendor.locality;
+
+	data = TPM_ACCESS_REQUEST_USE;
+	rc = I2C_WRITE_DATA(client, TPM_ACCESS, &data, 1);
+	if (rc < 0)
+		goto end;
+
+	if (chip->vendor.irq) {
+		rc = wait_for_serirq_timeout(chip, (check_locality
+						       (chip) >= 0),
+						      chip->vendor.timeout_a);
+		if (rc > 0)
+			return chip->vendor.locality;
+	} else{
+		stop = jiffies + chip->vendor.timeout_a;
+		do {
+			if (check_locality(chip) >= 0)
+				return chip->vendor.locality;
+			msleep(TPM_TIMEOUT);
+		} while (time_before(jiffies, stop));
+	}
+	rc = -EACCES;
+end:
+	return rc;
+} /* request_locality() */
+
+/*
+ * release_locality release the active locality
+ * @param: chip, the tpm chip description.
+ */
+static void release_locality(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	data = TPM_ACCESS_ACTIVE_LOCALITY;
+
+	I2C_WRITE_DATA(client, TPM_ACCESS, &data, 1);
+}
+
+/*
+ * get_burstcount return the burstcount address 0x19 0x1A
+ * @param: chip, the chip description
+ * return: the burstcount.
+ */
+static int get_burstcount(struct tpm_chip *chip)
+{
+	unsigned long stop;
+	int burstcnt, status;
+	u8 tpm_reg, temp;
+
+	struct i2c_client *client = (struct i2c_client *) chip->vendor.iobase;
+
+	stop = jiffies + chip->vendor.timeout_d;
+	do {
+		tpm_reg = TPM_STS + 1;
+		status = I2C_READ_DATA(client, tpm_reg, &temp, 1);
+		if (status < 0)
+			goto end;
+
+		tpm_reg = tpm_reg + 1;
+		burstcnt = temp;
+		status = I2C_READ_DATA(client, tpm_reg, &temp, 1);
+		if (status < 0)
+			goto end;
+
+		burstcnt |= temp << 8;
+		if (burstcnt)
+			return burstcnt;
+		msleep(TPM_TIMEOUT);
+	} while (time_before(jiffies, stop));
+
+end:
+	return -EBUSY;
+} /* get_burstcount() */
+
+/*
+ * wait_for_stat wait for a TPM_STS value
+ * @param: chip, the tpm chip description
+ * @param: mask, the value mask to wait
+ * @param: timeout, the timeout
+ * @param: queue, the wait queue.
+ * @return: the tpm status, 0 if success, -ETIME if timeout is reached.
+ */
+static int wait_for_stat(struct tpm_chip *chip, u8 mask, unsigned long timeout,
+			 wait_queue_head_t *queue)
+{
+	unsigned long stop;
+	long rc;
+	u8 status;
+
+	 if (chip->vendor.irq) {
+		rc = wait_for_serirq_timeout(chip, ((tpm_stm_i2c_status
+							(chip) & mask) ==
+						       mask), timeout);
+		if (rc > 0)
+			return 0;
+	} else{
+		stop = jiffies + timeout;
+		do {
+			msleep(TPM_TIMEOUT);
+			status = tpm_stm_i2c_status(chip);
+			if ((status & mask) == mask)
+				return 0;
+		} while (time_before(jiffies, stop));
+	}
+	return -ETIME;
+} /* wait_for_stat() */
+
+/*
+ * recv_data receive data
+ * @param: chip, the tpm chip description
+ * @param: buf, the buffer where the data are received
+ * @param: count, the number of data to receive
+ * @return: the number of bytes read from TPM FIFO.
+ */
+static int recv_data(struct tpm_chip *chip, u8 *buf, size_t count)
+{
+	int size = 0, burstcnt, len;
+	struct i2c_client *client;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	while (size < count &&
+	       wait_for_stat(chip,
+			     TPM_STS_DATA_AVAIL | TPM_STS_VALID,
+			     chip->vendor.timeout_c,
+			     &chip->vendor.read_queue)
+	       == 0) {
+		burstcnt = get_burstcount(chip);
+		len = min_t(int, burstcnt, count - size);
+		I2C_READ_DATA(client, TPM_DATA_FIFO, buf + size, len);
+		size += len;
+	}
+	return size;
+}
+
+/*
+ * tpm_ioserirq_handler the serirq irq handler
+ * @param: irq, the tpm chip description
+ * @param: dev_id, the description of the chip
+ * @return: the status of the handler.
+ */
+static irqreturn_t tpm_ioserirq_handler(int irq, void *dev_id)
+{
+	struct tpm_chip *chip = dev_id;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	disable_irq_nosync(irq);
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	complete(&pin_infos->irq_detection);
+	return IRQ_HANDLED;
+} /* tpm_ioserirq_handler() */
+
+
+/*
+ * tpm_stm_i2c_send send TPM commands through the I2C bus.
+ *
+ * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h
+ * @param: buf,	the buffer to send.
+ * @param: count, the number of bytes to send.
+ * @return: In case of success the number of bytes sent.
+ *			In other case, a < 0 value describing the issue.
+ */
+static int tpm_stm_i2c_send(struct tpm_chip *chip, unsigned char *buf,
+			    size_t len)
+{
+	u32 ret = 0, ordinal,
+	    status,
+	    burstcnt = 0, i, size;
+	u8 data;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	if (chip == NULL)
+		return -EBUSY;
+	if (len < TPM_HEADER_SIZE)
+		return -EBUSY;
+
+	client = (struct i2c_client *)chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	ordinal = be32_to_cpu(*((__be32 *) (buf + 6)));
+
+	client->flags = 0;
+
+	ret = request_locality(chip);
+	if (ret < 0)
+		return ret;
+
+	status = tpm_stm_i2c_status(chip);
+	if ((status & TPM_STS_COMMAND_READY) == 0) {
+		tpm_stm_i2c_cancel(chip);
+		if (wait_for_stat
+		    (chip, TPM_STS_COMMAND_READY, chip->vendor.timeout_b,
+		     &chip->vendor.int_queue) < 0) {
+			ret = -ETIME;
+			goto out_err;
+		}
+	}
+
+	for (i = 0 ; i < len - 1 ;) {
+		burstcnt = get_burstcount(chip);
+		size = min_t(int, len - i - 1, burstcnt);
+		ret = I2C_WRITE_DATA(client, TPM_DATA_FIFO, buf, size);
+		if (ret < 0)
+			goto out_err;
+
+		i += size;
+	}
+
+	status = tpm_stm_i2c_status(chip);
+	if ((status & TPM_STS_DATA_EXPECT) == 0) {
+		ret = -EIO;
+		goto out_err;
+	}
+
+	ret = I2C_WRITE_DATA(client, TPM_DATA_FIFO, buf + len - 1, 1);
+	if (ret < 0)
+		goto out_err;
+
+	status = tpm_stm_i2c_status(chip);
+	if ((status & TPM_STS_DATA_EXPECT) != 0) {
+		ret = -EIO;
+		goto out_err;
+	}
+
+	data = TPM_STS_GO;
+	I2C_WRITE_DATA(client, TPM_STS, &data, 1);
+
+	return len;
+out_err:
+	tpm_stm_i2c_cancel(chip);
+	release_locality(chip);
+	return ret;
+}
+
+/*
+ * tpm_stm_i2c_recv received TPM response through the I2C bus.
+ * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h.
+ * @param: buf,	the buffer to store datas.
+ * @param: count, the number of bytes to send.
+ * @return: In case of success the number of bytes received.
+ *		In other case, a < 0 value describing the issue.
+ */
+static int tpm_stm_i2c_recv(struct tpm_chip *chip, unsigned char *buf,
+			    size_t count)
+{
+	int size = 0;
+	int expected;
+
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	client = (struct i2c_client *)chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+
+	if (chip == NULL)
+		return -EBUSY;
+
+	if (count < TPM_HEADER_SIZE) {
+		size = -EIO;
+		goto out;
+	}
+
+	size = recv_data(chip, buf, TPM_HEADER_SIZE);
+	if (size < TPM_HEADER_SIZE) {
+		dev_err(chip->dev, "Unable to read header\n");
+		goto out;
+	}
+
+	expected = be32_to_cpu(*(__be32 *) (buf + 2));
+	if (expected > count) {
+		size = -EIO;
+		goto out;
+	}
+
+	size += recv_data(chip, &buf[TPM_HEADER_SIZE],
+					expected - TPM_HEADER_SIZE);
+	if (size < expected) {
+		dev_err(chip->dev, "Unable to read remainder of result\n");
+		size = -ETIME;
+		goto out;
+	}
+
+out:
+	chip->vendor.cancel(chip);
+	release_locality(chip);
+	return size;
+}
+
+static const struct file_operations tpm_st33_i2c_fops = {
+	.owner = THIS_MODULE,
+	.llseek = no_llseek,
+	.read = tpm_read,
+	.write = tpm_write,
+	.open = tpm_open,
+	.release = tpm_release,
+};
+
+static DEVICE_ATTR(pubek, S_IRUGO, tpm_show_pubek, NULL);
+static DEVICE_ATTR(pcrs, S_IRUGO, tpm_show_pcrs, NULL);
+static DEVICE_ATTR(enabled, S_IRUGO, tpm_show_enabled, NULL);
+static DEVICE_ATTR(active, S_IRUGO, tpm_show_active, NULL);
+static DEVICE_ATTR(owned, S_IRUGO, tpm_show_owned, NULL);
+static DEVICE_ATTR(temp_deactivated, S_IRUGO, tpm_show_temp_deactivated, NULL);
+static DEVICE_ATTR(caps, S_IRUGO, tpm_show_caps_1_2, NULL);
+static DEVICE_ATTR(cancel, S_IWUSR | S_IWGRP, NULL, tpm_store_cancel);
+
+static struct attribute *stm_tpm_attrs[] = {
+	&dev_attr_pubek.attr,
+	&dev_attr_pcrs.attr,
+	&dev_attr_enabled.attr,
+	&dev_attr_active.attr,
+	&dev_attr_owned.attr,
+	&dev_attr_temp_deactivated.attr,
+	&dev_attr_caps.attr,
+	&dev_attr_cancel.attr, NULL,
+};
+
+static struct attribute_group stm_tpm_attr_grp = {
+	.attrs = stm_tpm_attrs
+};
+
+static struct tpm_vendor_specific st_i2c_tpm = {
+	.send = tpm_stm_i2c_send,
+	.recv = tpm_stm_i2c_recv,
+	.cancel = tpm_stm_i2c_cancel,
+	.status = tpm_stm_i2c_status,
+	.req_complete_mask = TPM_STS_DATA_AVAIL | TPM_STS_VALID,
+	.req_complete_val = TPM_STS_DATA_AVAIL | TPM_STS_VALID,
+	.req_canceled = TPM_STS_COMMAND_READY,
+	.attr_group = &stm_tpm_attr_grp,
+	.miscdev = {.fops = &tpm_st33_i2c_fops,},
+};
+
+static int interrupts ;
+module_param(interrupts, int, 0444);
+MODULE_PARM_DESC(interrupts, "Enable interrupts");
+
+static int power_mgt = 1;
+module_param(power_mgt, int, 0444);
+MODULE_PARM_DESC(power_mgt, "Power Management");
+
+/*
+ * tpm_st33_i2c_probe initialize the TPM device
+ * @param: client, the i2c_client drescription (TPM I2C description).
+ * @param: id, the i2c_device_id struct.
+ * @return: 0 in case of success.
+ *	 -1 in other case.
+ */
+static int
+tpm_st33_i2c_probe(struct i2c_client *client, const struct i2c_device_id *id)
+{
+	u32 err;
+	u8 intmask;
+	struct tpm_chip *chip;
+	struct st33zp24_platform_data *platform_data;
+
+	err = 0;
+
+	if (client == NULL) {
+		dev_info(&client->dev, "client is NULL. exiting.\n");
+		err = -ENODEV;
+		goto end;
+	}
+
+	if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
+		dev_info(&client->dev, "client not i2c capable\n");
+		err = -ENODEV;
+		goto end;
+	}
+
+	chip = tpm_register_hardware(&client->dev, &st_i2c_tpm);
+	if (!chip) {
+		dev_info(&client->dev, "fail chip\n");
+		err = -ENODEV;
+		goto end;
+	}
+
+	platform_data = client->dev.platform_data;
+	platform_data->tpm_i2c_buffer[0] =
+	    kmalloc(TPM_BUFSIZE * sizeof(u8), GFP_KERNEL);
+	if (platform_data->tpm_i2c_buffer[0] == NULL) {
+		err = -ENOMEM;
+		goto _tpm_clean_answer;
+	}
+	platform_data->tpm_i2c_buffer[1] =
+	    kmalloc(TPM_BUFSIZE * sizeof(u8), GFP_KERNEL);
+	if (platform_data->tpm_i2c_buffer[1] == NULL) {
+		err = -ENOMEM;
+		goto _tpm_clean_response;
+	}
+
+	chip->vendor.iobase = client;
+
+	chip->vendor.timeout_a = msecs_to_jiffies(TIS_SHORT_TIMEOUT);
+	chip->vendor.timeout_b = msecs_to_jiffies(TIS_LONG_TIMEOUT);
+	chip->vendor.timeout_c = msecs_to_jiffies(TIS_SHORT_TIMEOUT);
+	chip->vendor.timeout_d = msecs_to_jiffies(TIS_SHORT_TIMEOUT);
+
+	chip->vendor.locality = LOCALITY0;
+
+	if (power_mgt) {
+		err = gpio_request(platform_data->io_lpcpd, "TPM IO_LPCPD");
+		if (err)
+			goto _gpio_init1;
+		gpio_set_value(platform_data->io_lpcpd, 1);
+	}
+
+	if (interrupts) {
+		init_completion(&platform_data->irq_detection);
+		if (request_locality(chip) != LOCALITY0) {
+			err = -ENODEV;
+			goto _tpm_clean_response;
+		}
+		err = gpio_request(platform_data->io_serirq, "TPM IO_SERIRQ");
+		if (err)
+			goto _gpio_init2;
+
+		clear_interruption(client);
+		err = request_irq(gpio_to_irq(platform_data->io_serirq),
+				&tpm_ioserirq_handler,
+				IRQF_TRIGGER_HIGH,
+				"TPM SERIRQ management", chip);
+		if (err < 0) {
+			dev_err(chip->dev , "TPM SERIRQ signals %d not available\n",
+			gpio_to_irq(platform_data->io_serirq));
+			goto _irq_set;
+		}
+
+		err = I2C_READ_DATA(client, TPM_INT_ENABLE, &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		intmask |= TPM_INTF_CMD_READY_INT
+			|  TPM_INTF_FIFO_AVALAIBLE_INT
+			|  TPM_INTF_WAKE_UP_READY_INT
+			|  TPM_INTF_LOCALITY_CHANGE_INT
+			|  TPM_INTF_STS_VALID_INT
+			|  TPM_INTF_DATA_AVAIL_INT;
+
+		err = I2C_WRITE_DATA(client, TPM_INT_ENABLE, &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		intmask = TPM_GLOBAL_INT_ENABLE;
+		err = I2C_WRITE_DATA(client, TPM_INT_ENABLE + 3, &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		err = I2C_READ_DATA(client, TPM_INT_STATUS, &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		chip->vendor.irq = interrupts;
+
+		tpm_gen_interrupt(chip);
+	}
+
+	tpm_get_timeouts(chip);
+
+	i2c_set_clientdata(client, chip);
+	platform_data->bChipF = false;
+
+	dev_info(chip->dev, "TPM I2C Initialized\n");
+	return 0;
+_irq_set:
+	free_irq(gpio_to_irq(platform_data->io_serirq), (void *) chip);
+_gpio_init2:
+	if (platform_data && interrupts)
+		gpio_free(platform_data->io_serirq);
+_gpio_init1:
+	if (platform_data && power_mgt)
+		gpio_free(platform_data->io_lpcpd);
+_tpm_clean_response:
+	tpm_remove_hardware(chip->dev);
+	if (platform_data->tpm_i2c_buffer[1] != NULL) {
+		kzfree(platform_data->tpm_i2c_buffer[1]);
+		platform_data->tpm_i2c_buffer[1] = NULL;
+	}
+_tpm_clean_answer:
+	if (platform_data->tpm_i2c_buffer[0] != NULL) {
+		kzfree(platform_data->tpm_i2c_buffer[0]);
+		platform_data->tpm_i2c_buffer[0] = NULL;
+	}
+
+	platform_data->bChipF = true;
+end:
+	pr_info("TPM I2C initialisation fail\n");
+	return err;
+}
+
+/*
+ * tpm_st33_i2c_remove remove the TPM device
+ * @param: client, the i2c_client drescription (TPM I2C description).
+		clear_bit(0, &chip->is_open);
+ * @return: 0 in case of success.
+ */
+static __devexit int tpm_st33_i2c_remove(struct i2c_client *client)
+{
+	struct tpm_chip *chip = (struct tpm_chip *)i2c_get_clientdata(client);
+	struct st33zp24_platform_data *pin_infos =
+		((struct i2c_client *) chip->vendor.iobase)->dev.platform_data;
+
+	if (pin_infos != NULL) {
+		free_irq(pin_infos->io_serirq, chip);
+
+		gpio_free(pin_infos->io_serirq);
+		gpio_free(pin_infos->io_lpcpd);
+
+		if (pin_infos->bChipF != true)
+			tpm_remove_hardware(chip->dev);
+		if (pin_infos->tpm_i2c_buffer[1] != NULL) {
+			kzfree(pin_infos->tpm_i2c_buffer[1]);
+			pin_infos->tpm_i2c_buffer[1] = NULL;
+		}
+		if (pin_infos->tpm_i2c_buffer[0] != NULL) {
+			kzfree(pin_infos->tpm_i2c_buffer[0]);
+			pin_infos->tpm_i2c_buffer[0] = NULL;
+		}
+	}
+
+	return 0;
+}
+
+/*
+ * tpm_st33_i2c_pm_suspend suspend the TPM device
+ * Added: Work around when suspend and no tpm application is running, suspend
+ * may fail because chip->data_buffer is not set (only set in tpm_open in Linux
+ * TPM core)
+ * @param: client, the i2c_client drescription (TPM I2C description).
+ * @param: mesg, the power management message.
+ * @return: 0 in case of success.
+ */
+static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
+{
+	struct tpm_chip *chip =
+	    (struct tpm_chip *)i2c_get_clientdata(client);
+	struct st33zp24_platform_data *pin_infos =
+		((struct i2c_client *)chip->vendor.iobase)->dev.platform_data;
+	int ret = 0;
+
+	if (power_mgt)
+		gpio_set_value(pin_infos->io_lpcpd, 0);
+	else{
+		if (chip->data_buffer == NULL)
+			chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
+		ret = tpm_pm_suspend(&client->dev, mesg);
+	}
+	return ret;
+}				/* tpm_st33_i2c_suspend() */
+
+/*
+ * tpm_st33_i2c_pm_resume resume the TPM device
+ * @param: client, the i2c_client drescription (TPM I2C description).
+ * @return: 0 in case of success.
+ */
+static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
+{
+	struct tpm_chip *chip =
+	    (struct tpm_chip *)i2c_get_clientdata(client);
+	struct st33zp24_platform_data *pin_infos =
+		((struct i2c_client *)chip->vendor.iobase)->dev.platform_data;
+
+	int ret = 0;
+
+	if (power_mgt) {
+		gpio_set_value(pin_infos->io_lpcpd, 1);
+		ret = wait_for_serirq_timeout(chip,
+					  (chip->vendor.status(chip) &&
+					  TPM_STS_VALID) == TPM_STS_VALID,
+					  chip->vendor.timeout_b);
+	} else{
+	if (chip->data_buffer == NULL)
+		chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
+	ret = tpm_pm_resume(&client->dev);
+	if (!ret)
+		tpm_continue_selftest_nocheck(chip);
+	}
+	return ret;
+}				/* tpm_st33_i2c_pm_resume() */
+
+static const struct i2c_device_id tpm_st33_i2c_id[] = {
+	{TPM_ST33_I2C, 0},
+	{}
+};
+
+MODULE_DEVICE_TABLE(i2c, tpm_st33_i2c_id);
+
+static struct i2c_driver tpm_st33_i2c_driver = {
+	.driver = {
+		   .owner = THIS_MODULE,
+		   .name = TPM_ST33_I2C,
+		   },
+	.probe = tpm_st33_i2c_probe,
+	.remove = tpm_st33_i2c_remove,
+	.resume = tpm_st33_i2c_pm_resume,
+	.suspend = tpm_st33_i2c_pm_suspend,
+	.id_table = tpm_st33_i2c_id
+};
+
+/*
+ * tpm_st33_i2c_init initialize driver
+ * @return: 0 if successful, else non zero value.
+ */
+static int __init tpm_st33_i2c_init(void)
+{
+	return i2c_add_driver(&tpm_st33_i2c_driver);
+}
+
+/*
+ * tpm_st33_i2c_exit The kernel calls this function during unloading the
+ * module or during shut down process
+ */
+static void __exit tpm_st33_i2c_exit(void)
+{
+	i2c_del_driver(&tpm_st33_i2c_driver);
+}
+
+module_init(tpm_st33_i2c_init);
+module_exit(tpm_st33_i2c_exit);
+
+MODULE_AUTHOR("Christophe Ricard ([email protected])");
+MODULE_DESCRIPTION("STM TPM I2C ST33 Driver");
+MODULE_VERSION("1.2.0");

+MODULE_LICENSE("GPL");

drivers/char/tpm/tpm_stm_st33_i2c.h new file +81 -0

81 lines not on this page · raw message

include/linux/i2c/tpm_stm_st33_i2c.h new file +47 -0

47 lines not on this page · raw message

Signature · 1 line
1.7.9.5
#1 lore raw

PATCH 1/1 TPM: STMicroelectronics ST33 I2C KERNEL 2.6

* STMicroelectronics version 1.2.0, Copyright (C) 2010
* STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
* This is free software, and you are welcome to redistribute it
* under certain conditions.

This is the driver for TPM chip from ST Microelectronics.

If you have a TPM security chip from STMicroelectronics working with an I2C, in menuconfig or .config choose the tpm driver on device --> tpm and activate the protocol of your choice before compiling the kernel.
The driver will be accessible from within Linux.

Tested on linux x86/x64, beagleboard REV B & XM REV C and CHROMIUM OS on kernel 2.6

---
 drivers/char/tpm/tpm_stm_st33_i2c.c  |  904 ++++++++++++++++++++++++++++++++++
 drivers/char/tpm/tpm_stm_st33_i2c.h  |   81 +++
 include/linux/i2c/tpm_stm_st33_i2c.h |   47 ++
 3 files changed, 1032 insertions(+)
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.c
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.h
 create mode 100644 include/linux/i2c/tpm_stm_st33_i2c.h
drivers/char/tpm/tpm_stm_st33_i2c.c new file +904 -0
new file mode 100644
index 0000000..bf18885
@@ -0,0 +1,904 @@
+/*
+ * STMicroelectronics TPM I2C Linux driver for TPM ST33ZP24
+ * Copyright (C) 2009, 2010  STMicroelectronics
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * STMicroelectronics version 1.2.0, Copyright (C) 2010
+ * STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
+ * This is free software, and you are welcome to redistribute it
+ * under certain conditions.
+ *
+ * @Author: Christophe RICARD [email protected]
+ *
+ * @File: tpm_stm_st33_i2c.c
+ *
+ * @Synopsis:
+ *	09/15/2010:	First shot driver tpm_tis driver for
+			 lpc is used as model.
+ */
+
+#include <linux/i2c/tpm_stm_st33_i2c.h>
+
+#include "tpm_stm_st33_i2c.h"
+
+enum stm33zp24_access {
+	TPM_ACCESS_VALID = 0x80,
+	TPM_ACCESS_ACTIVE_LOCALITY = 0x20,
+	TPM_ACCESS_REQUEST_PENDING = 0x04,
+	TPM_ACCESS_REQUEST_USE = 0x02,
+};
+
+enum stm33zp24_status {
+	TPM_STS_VALID = 0x80,
+	TPM_STS_COMMAND_READY = 0x40,
+	TPM_STS_GO = 0x20,
+	TPM_STS_DATA_AVAIL = 0x10,
+	TPM_STS_DATA_EXPECT = 0x08,
+};
+
+enum stm33zp24_int_flags {
+	TPM_GLOBAL_INT_ENABLE = 0x80,
+	TPM_INTF_CMD_READY_INT = 0x080,
+	TPM_INTF_FIFO_AVALAIBLE_INT = 0x040,
+	TPM_INTF_WAKE_UP_READY_INT = 0x020,
+	TPM_INTF_LOCALITY_CHANGE_INT = 0x004,
+	TPM_INTF_STS_VALID_INT = 0x002,
+	TPM_INTF_DATA_AVAIL_INT = 0x001,
+};
+
+enum tis_defaults {
+	TIS_SHORT_TIMEOUT = 750,
+	TIS_LONG_TIMEOUT = 2000,
+};
+
+/*
+ * write8_reg
+ * Send byte to the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: tpm_register, the tpm tis register where the data should be written
+ * @param: tpm_data, the tpm_data to write inside the tpm_register
+ * @param: tpm_size, The length of the data
+ * @return: Returns negative errno, or else the number of bytes written.
+ */
+static int write8_reg(struct i2c_client *client, u8 tpm_register,
+		      u8 *tpm_data, u16 tpm_size)
+{
+	u8 data;
+	int value = 0;
+	struct st33zp24_platform_data *pin_infos;
+
+	pin_infos = client->dev.platform_data;
+
+	data = tpm_register;
+	memcpy(pin_infos->tpm_i2c_buffer[0], &data, sizeof(data));
+	memcpy(pin_infos->tpm_i2c_buffer[0] + 1, tpm_data, tpm_size);
+	value = i2c_master_send(client, pin_infos->tpm_i2c_buffer[0],
+				tpm_size + 1);
+	return value;
+} /* write8_reg() */
+
+/*
+ * read8_reg
+ * Recv byte from the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: tpm_register, the tpm tis register where the data should be read
+ * @param: tpm_data, the TPM response
+ * @param: tpm_size, tpm TPM response size to read.
+ * @return: number of byte read successfully: should be one if success.
+ */
+static int read8_reg(struct i2c_client *client, u8 tpm_register,
+		    u8 *tpm_data, int tpm_size)
+{
+	u8 status = 0;
+	u8 data;
+	struct st33zp24_platform_data *pin_infos;
+
+	pin_infos = client->dev.platform_data;
+
+	data = TPM_DUMMY_BYTE;
+	status = write8_reg(client, tpm_register, &data, 1);
+	if (status == 2)
+		status = i2c_master_recv(client, tpm_data, tpm_size);
+	return status;
+} /* read8_reg() */
+
+/*
+ * I2C_WRITE_DATA
+ * Send byte to the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: client, the chip description
+ * @param: tpm_register, the tpm tis register where the data should be written
+ * @param: tpm_data, the tpm_data to write inside the tpm_register
+ * @param: tpm_size, The length of the data
+ * @return: number of byte written successfully: should be one if success.
+ */
+#define I2C_WRITE_DATA(client, tpm_register, tpm_data, tpm_size) \
+	(write8_reg(client, tpm_register | \
+	TPM_WRITE_DIRECTION, tpm_data, tpm_size))
+
+/*
+ * I2C_READ_DATA
+ * Recv byte from the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: tpm, the chip description
+ * @param: tpm_register, the tpm tis register where the data should be read
+ * @param: tpm_data, the TPM response
+ * @param: tpm_size, tpm TPM response size to read.
+ * @return: number of byte read successfully: should be one if success.
+ */
+#define I2C_READ_DATA(client, tpm_register, tpm_data, tpm_size) \
+	(read8_reg(client, tpm_register, tpm_data, tpm_size))
+
+/*
+ * clear_interruption
+ * clear the TPM interrupt register.
+ * @param: tpm, the chip description
+ */
+static void clear_interruption(struct i2c_client *client)
+{
+	u8 interrupt;
+	I2C_READ_DATA(client, TPM_INT_STATUS, &interrupt, 1);
+	I2C_WRITE_DATA(client, TPM_INT_STATUS, &interrupt, 1);
+	I2C_READ_DATA(client, TPM_INT_STATUS, &interrupt, 1);
+} /* clear_interruption() */
+
+/*
+ * _wait_for_interrupt_serirq_timeout
+ * @param: tpm, the chip description
+ * @param: timeout, the timeout of the interrupt
+ * @return: the status of the interruption.
+ */
+static long _wait_for_interrupt_serirq_timeout(struct tpm_chip *chip,
+						unsigned long timeout)
+{
+	long status;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	status = wait_for_completion_interruptible_timeout(
+					&pin_infos->irq_detection,
+						timeout);
+	if (status > 0)
+		enable_irq(gpio_to_irq(pin_infos->io_serirq));
+	gpio_direction_input(pin_infos->io_serirq);
+
+	return status;
+} /* wait_for_interrupt_serirq_timeout() */
+
+int wait_for_serirq_timeout(struct tpm_chip *chip, bool condition,
+				 unsigned long timeout)
+{
+	int status = 2;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	status = _wait_for_interrupt_serirq_timeout(chip, timeout);
+	if (!status) {
+		status = -EBUSY;
+	} else{
+		clear_interruption(client);
+		if (condition)
+			status = 1;
+	}
+	return status;
+}
+
+/*
+ * tpm_stm_i2c_cancel, cancel is not implemented.
+ * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h
+ */
+static void tpm_stm_i2c_cancel(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	data = TPM_STS_COMMAND_READY;
+	I2C_WRITE_DATA(client, TPM_STS, &data, 1);
+	if (chip->vendor.irq)
+		wait_for_serirq_timeout(chip, 1, chip->vendor.timeout_a);
+}	/* tpm_stm_i2c_cancel() */
+
+/*
+ * tpm_stm_spi_status return the TPM_STS register
+ * @param: chip, the tpm chip description
+ * @return: the TPM_STS register value.
+ */
+static u8 tpm_stm_i2c_status(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	I2C_READ_DATA(client, TPM_STS, &data, 1);
+	return data;
+}				/* tpm_stm_i2c_status() */
+
+
+/*
+ * check_locality if the locality is active
+ * @param: chip, the tpm chip description
+ * @return: the active locality or -EACCESS.
+ */
+static int check_locality(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+	u8 status;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	status = I2C_READ_DATA(client, TPM_ACCESS, &data, 1);
+	if (status && (data &
+		(TPM_ACCESS_ACTIVE_LOCALITY | TPM_ACCESS_VALID)) ==
+		(TPM_ACCESS_ACTIVE_LOCALITY | TPM_ACCESS_VALID))
+		return chip->vendor.locality;
+
+	return -EACCES;
+
+} /* check_locality() */
+
+/*
+ * request_locality request the TPM locality
+ * @param: chip, the chip description
+ * @return: the active locality or EACCESS.
+ */
+static int request_locality(struct tpm_chip *chip)
+{
+	unsigned long stop;
+	long rc;
+	struct i2c_client *client;
+	u8 data;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	if (check_locality(chip) == chip->vendor.locality)
+		return chip->vendor.locality;
+
+	data = TPM_ACCESS_REQUEST_USE;
+	rc = I2C_WRITE_DATA(client, TPM_ACCESS, &data, 1);
+	if (rc < 0)
+		goto end;
+
+	if (chip->vendor.irq) {
+		rc = wait_for_serirq_timeout(chip, (check_locality
+						       (chip) >= 0),
+						      chip->vendor.timeout_a);
+		if (rc > 0)
+			return chip->vendor.locality;
+	} else{
+		stop = jiffies + chip->vendor.timeout_a;
+		do {
+			if (check_locality(chip) >= 0)
+				return chip->vendor.locality;
+			msleep(TPM_TIMEOUT);
+		} while (time_before(jiffies, stop));
+	}
+	rc = -EACCES;
+end:
+	return rc;
+} /* request_locality() */
+
+/*
+ * release_locality release the active locality
+ * @param: chip, the tpm chip description.
+ */
+static void release_locality(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	data = TPM_ACCESS_ACTIVE_LOCALITY;
+
+	I2C_WRITE_DATA(client, TPM_ACCESS, &data, 1);
+}
+
+/*
+ * get_burstcount return the burstcount address 0x19 0x1A
+ * @param: chip, the chip description
+ * return: the burstcount.
+ */
+static int get_burstcount(struct tpm_chip *chip)
+{
+	unsigned long stop;
+	int burstcnt, status;
+	u8 tpm_reg, temp;
+
+	struct i2c_client *client = (struct i2c_client *) chip->vendor.iobase;
+
+	stop = jiffies + chip->vendor.timeout_d;
+	do {
+		tpm_reg = TPM_STS + 1;
+		status = I2C_READ_DATA(client, tpm_reg, &temp, 1);
+		if (status < 0)
+			goto end;
+
+		tpm_reg = tpm_reg + 1;
+		burstcnt = temp;
+		status = I2C_READ_DATA(client, tpm_reg, &temp, 1);
+		if (status < 0)
+			goto end;
+
+		burstcnt |= temp << 8;
+		if (burstcnt)
+			return burstcnt;
+		msleep(TPM_TIMEOUT);
+	} while (time_before(jiffies, stop));
+
+end:
+	return -EBUSY;
+} /* get_burstcount() */
+
+/*
+ * wait_for_stat wait for a TPM_STS value
+ * @param: chip, the tpm chip description
+ * @param: mask, the value mask to wait
+ * @param: timeout, the timeout
+ * @param: queue, the wait queue.
+ * @return: the tpm status, 0 if success, -ETIME if timeout is reached.
+ */
+static int wait_for_stat(struct tpm_chip *chip, u8 mask, unsigned long timeout,
+			 wait_queue_head_t *queue)
+{
+	unsigned long stop;
+	long rc;
+	u8 status;
+
+	 if (chip->vendor.irq) {
+		rc = wait_for_serirq_timeout(chip, ((tpm_stm_i2c_status
+							(chip) & mask) ==
+						       mask), timeout);
+		if (rc > 0)
+			return 0;
+	} else{
+		stop = jiffies + timeout;
+		do {
+			msleep(TPM_TIMEOUT);
+			status = tpm_stm_i2c_status(chip);
+			if ((status & mask) == mask)
+				return 0;
+		} while (time_before(jiffies, stop));
+	}
+	return -ETIME;
+} /* wait_for_stat() */
+
+/*
+ * recv_data receive data
+ * @param: chip, the tpm chip description
+ * @param: buf, the buffer where the data are received
+ * @param: count, the number of data to receive
+ * @return: the number of bytes read from TPM FIFO.
+ */
+static int recv_data(struct tpm_chip *chip, u8 *buf, size_t count)
+{
+	int size = 0, burstcnt, len;
+	struct i2c_client *client;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	while (size < count &&
+	       wait_for_stat(chip,
+			     TPM_STS_DATA_AVAIL | TPM_STS_VALID,
+			     chip->vendor.timeout_c,
+			     &chip->vendor.read_queue)
+	       == 0) {
+		burstcnt = get_burstcount(chip);
+		len = min_t(int, burstcnt, count - size);
+		I2C_READ_DATA(client, TPM_DATA_FIFO, buf + size, len);
+		size += len;
+	}
+	return size;
+}
+
+/*
+ * tpm_ioserirq_handler the serirq irq handler
+ * @param: irq, the tpm chip description
+ * @param: dev_id, the description of the chip
+ * @return: the status of the handler.
+ */
+static irqreturn_t tpm_ioserirq_handler(int irq, void *dev_id)
+{
+	struct tpm_chip *chip = dev_id;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	disable_irq_nosync(irq);
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	complete(&pin_infos->irq_detection);
+	return IRQ_HANDLED;
+} /* tpm_ioserirq_handler() */
+
+
+/*
+ * tpm_stm_i2c_send send TPM commands through the I2C bus.
+ *
+ * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h
+ * @param: buf,	the buffer to send.
+ * @param: count, the number of bytes to send.
+ * @return: In case of success the number of bytes sent.
+ *			In other case, a < 0 value describing the issue.
+ */
+static int tpm_stm_i2c_send(struct tpm_chip *chip, unsigned char *buf,
+			    size_t len)
+{
+	u32 ret = 0, ordinal,
+	    status,
+	    burstcnt = 0, i, size;
+	u8 data;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	if (chip == NULL)
+		return -EBUSY;
+	if (len < TPM_HEADER_SIZE)
+		return -EBUSY;
+
+	client = (struct i2c_client *)chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	ordinal = be32_to_cpu(*((__be32 *) (buf + 6)));
+
+	client->flags = 0;
+
+	ret = request_locality(chip);
+	if (ret < 0)
+		return ret;
+
+	status = tpm_stm_i2c_status(chip);
+	if ((status & TPM_STS_COMMAND_READY) == 0) {
+		tpm_stm_i2c_cancel(chip);
+		if (wait_for_stat
+		    (chip, TPM_STS_COMMAND_READY, chip->vendor.timeout_b,
+		     &chip->vendor.int_queue) < 0) {
+			ret = -ETIME;
+			goto out_err;
+		}
+	}
+
+	for (i = 0 ; i < len - 1 ;) {
+		burstcnt = get_burstcount(chip);
+		size = min_t(int, len - i - 1, burstcnt);
+		ret = I2C_WRITE_DATA(client, TPM_DATA_FIFO, buf, size);
+		if (ret < 0)
+			goto out_err;
+
+		i += size;
+	}
+
+	status = tpm_stm_i2c_status(chip);
+	if ((status & TPM_STS_DATA_EXPECT) == 0) {
+		ret = -EIO;
+		goto out_err;
+	}
+
+	ret = I2C_WRITE_DATA(client, TPM_DATA_FIFO, buf + len - 1, 1);
+	if (ret < 0)
+		goto out_err;
+
+	status = tpm_stm_i2c_status(chip);
+	if ((status & TPM_STS_DATA_EXPECT) != 0) {
+		ret = -EIO;
+		goto out_err;
+	}
+
+	data = TPM_STS_GO;
+	I2C_WRITE_DATA(client, TPM_STS, &data, 1);
+
+	return len;
+out_err:
+	tpm_stm_i2c_cancel(chip);
+	release_locality(chip);
+	return ret;
+}
+
+/*
+ * tpm_stm_i2c_recv received TPM response through the I2C bus.
+ * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h.
+ * @param: buf,	the buffer to store datas.
+ * @param: count, the number of bytes to send.
+ * @return: In case of success the number of bytes received.
+ *		In other case, a < 0 value describing the issue.
+ */
+static int tpm_stm_i2c_recv(struct tpm_chip *chip, unsigned char *buf,
+			    size_t count)
+{
+	int size = 0;
+	int expected;
+
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	client = (struct i2c_client *)chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+
+	if (chip == NULL)
+		return -EBUSY;
+
+	if (count < TPM_HEADER_SIZE) {
+		size = -EIO;
+		goto out;
+	}
+
+	size = recv_data(chip, buf, TPM_HEADER_SIZE);
+	if (size < TPM_HEADER_SIZE) {
+		dev_err(chip->dev, "Unable to read header\n");
+		goto out;
+	}
+
+	expected = be32_to_cpu(*(__be32 *) (buf + 2));
+	if (expected > count) {
+		size = -EIO;
+		goto out;
+	}
+
+	size += recv_data(chip, &buf[TPM_HEADER_SIZE],
+					expected - TPM_HEADER_SIZE);
+	if (size < expected) {
+		dev_err(chip->dev, "Unable to read remainder of result\n");
+		size = -ETIME;
+		goto out;
+	}
+
+out:
+	chip->vendor.cancel(chip);
+	release_locality(chip);
+	return size;
+}
+
+static const struct file_operations tpm_st33_i2c_fops = {
+	.owner = THIS_MODULE,
+	.llseek = no_llseek,
+	.read = tpm_read,
+	.write = tpm_write,
+	.open = tpm_open,
+	.release = tpm_release,
+};
+
+static DEVICE_ATTR(pubek, S_IRUGO, tpm_show_pubek, NULL);
+static DEVICE_ATTR(pcrs, S_IRUGO, tpm_show_pcrs, NULL);
+static DEVICE_ATTR(enabled, S_IRUGO, tpm_show_enabled, NULL);
+static DEVICE_ATTR(active, S_IRUGO, tpm_show_active, NULL);
+static DEVICE_ATTR(owned, S_IRUGO, tpm_show_owned, NULL);
+static DEVICE_ATTR(temp_deactivated, S_IRUGO, tpm_show_temp_deactivated, NULL);
+static DEVICE_ATTR(caps, S_IRUGO, tpm_show_caps_1_2, NULL);
+static DEVICE_ATTR(cancel, S_IWUSR | S_IWGRP, NULL, tpm_store_cancel);
+
+static struct attribute *stm_tpm_attrs[] = {
+	&dev_attr_pubek.attr,
+	&dev_attr_pcrs.attr,
+	&dev_attr_enabled.attr,
+	&dev_attr_active.attr,
+	&dev_attr_owned.attr,
+	&dev_attr_temp_deactivated.attr,
+	&dev_attr_caps.attr,
+	&dev_attr_cancel.attr, NULL,
+};
+
+static struct attribute_group stm_tpm_attr_grp = {
+	.attrs = stm_tpm_attrs
+};
+
+static struct tpm_vendor_specific st_i2c_tpm = {
+	.send = tpm_stm_i2c_send,
+	.recv = tpm_stm_i2c_recv,
+	.cancel = tpm_stm_i2c_cancel,
+	.status = tpm_stm_i2c_status,
+	.req_complete_mask = TPM_STS_DATA_AVAIL | TPM_STS_VALID,
+	.req_complete_val = TPM_STS_DATA_AVAIL | TPM_STS_VALID,
+	.req_canceled = TPM_STS_COMMAND_READY,
+	.attr_group = &stm_tpm_attr_grp,
+	.miscdev = {.fops = &tpm_st33_i2c_fops,},
+};
+
+static int interrupts ;
+module_param(interrupts, int, 0444);
+MODULE_PARM_DESC(interrupts, "Enable interrupts");
+
+static int power_mgt = 1;
+module_param(power_mgt, int, 0444);
+MODULE_PARM_DESC(power_mgt, "Power Management");
+
+/*
+ * tpm_st33_i2c_probe initialize the TPM device
+ * @param: client, the i2c_client drescription (TPM I2C description).
+ * @param: id, the i2c_device_id struct.
+ * @return: 0 in case of success.
+ *	 -1 in other case.
+ */
+static int
+tpm_st33_i2c_probe(struct i2c_client *client, const struct i2c_device_id *id)
+{
+	u32 err;
+	u8 intmask;
+	struct tpm_chip *chip;
+	struct st33zp24_platform_data *platform_data;
+
+	err = 0;
+
+	if (client == NULL) {
+		dev_info(&client->dev, "client is NULL. exiting.\n");
+		err = -ENODEV;
+		goto end;
+	}
+
+	if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
+		dev_info(&client->dev, "client not i2c capable\n");
+		err = -ENODEV;
+		goto end;
+	}
+
+	chip = tpm_register_hardware(&client->dev, &st_i2c_tpm);
+	if (!chip) {
+		dev_info(&client->dev, "fail chip\n");
+		err = -ENODEV;
+		goto end;
+	}
+
+	platform_data = client->dev.platform_data;
+	platform_data->tpm_i2c_buffer[0] =
+	    kmalloc(TPM_BUFSIZE * sizeof(u8), GFP_KERNEL);
+	if (platform_data->tpm_i2c_buffer[0] == NULL) {
+		err = -ENOMEM;
+		goto _tpm_clean_answer;
+	}
+	platform_data->tpm_i2c_buffer[1] =
+	    kmalloc(TPM_BUFSIZE * sizeof(u8), GFP_KERNEL);
+	if (platform_data->tpm_i2c_buffer[1] == NULL) {
+		err = -ENOMEM;
+		goto _tpm_clean_response;
+	}
+
+	chip->vendor.iobase = client;
+
+	chip->vendor.timeout_a = msecs_to_jiffies(TIS_SHORT_TIMEOUT);
+	chip->vendor.timeout_b = msecs_to_jiffies(TIS_LONG_TIMEOUT);
+	chip->vendor.timeout_c = msecs_to_jiffies(TIS_SHORT_TIMEOUT);
+	chip->vendor.timeout_d = msecs_to_jiffies(TIS_SHORT_TIMEOUT);
+
+	chip->vendor.locality = LOCALITY0;
+
+	if (power_mgt) {
+		err = gpio_request(platform_data->io_lpcpd, "TPM IO_LPCPD");
+		if (err)
+			goto _gpio_init1;
+		gpio_set_value(platform_data->io_lpcpd, 1);
+	}
+
+	if (interrupts) {
+		init_completion(&platform_data->irq_detection);
+		if (request_locality(chip) != LOCALITY0) {
+			err = -ENODEV;
+			goto _tpm_clean_response;
+		}
+		err = gpio_request(platform_data->io_serirq, "TPM IO_SERIRQ");
+		if (err)
+			goto _gpio_init2;
+
+		clear_interruption(client);
+		err = request_irq(gpio_to_irq(platform_data->io_serirq),
+				&tpm_ioserirq_handler,
+				IRQF_TRIGGER_HIGH,
+				"TPM SERIRQ management", chip);
+		if (err < 0) {
+			dev_err(chip->dev , "TPM SERIRQ signals %d not available\n",
+			gpio_to_irq(platform_data->io_serirq));
+			goto _irq_set;
+		}
+
+		err = I2C_READ_DATA(client, TPM_INT_ENABLE, &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		intmask |= TPM_INTF_CMD_READY_INT
+			|  TPM_INTF_FIFO_AVALAIBLE_INT
+			|  TPM_INTF_WAKE_UP_READY_INT
+			|  TPM_INTF_LOCALITY_CHANGE_INT
+			|  TPM_INTF_STS_VALID_INT
+			|  TPM_INTF_DATA_AVAIL_INT;
+
+		err = I2C_WRITE_DATA(client, TPM_INT_ENABLE, &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		intmask = TPM_GLOBAL_INT_ENABLE;
+		err = I2C_WRITE_DATA(client, TPM_INT_ENABLE + 3, &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		err = I2C_READ_DATA(client, TPM_INT_STATUS, &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		chip->vendor.irq = interrupts;
+
+		tpm_gen_interrupt(chip);
+	}
+
+	tpm_get_timeouts(chip);
+
+	i2c_set_clientdata(client, chip);
+	platform_data->bChipF = false;
+
+	dev_info(chip->dev, "TPM I2C Initialized\n");
+	return 0;
+_irq_set:
+	free_irq(gpio_to_irq(platform_data->io_serirq), (void *) chip);
+_gpio_init2:
+	if (platform_data && interrupts)
+		gpio_free(platform_data->io_serirq);
+_gpio_init1:
+	if (platform_data && power_mgt)
+		gpio_free(platform_data->io_lpcpd);
+_tpm_clean_response:
+	tpm_remove_hardware(chip->dev);
+	if (platform_data->tpm_i2c_buffer[1] != NULL) {
+		kzfree(platform_data->tpm_i2c_buffer[1]);
+		platform_data->tpm_i2c_buffer[1] = NULL;
+	}
+_tpm_clean_answer:
+	if (platform_data->tpm_i2c_buffer[0] != NULL) {
+		kzfree(platform_data->tpm_i2c_buffer[0]);
+		platform_data->tpm_i2c_buffer[0] = NULL;
+	}
+
+	platform_data->bChipF = true;
+end:
+	pr_info("TPM I2C initialisation fail\n");
+	return err;
+}
+
+/*
+ * tpm_st33_i2c_remove remove the TPM device
+ * @param: client, the i2c_client drescription (TPM I2C description).
+		clear_bit(0, &chip->is_open);
+ * @return: 0 in case of success.
+ */
+static __devexit int tpm_st33_i2c_remove(struct i2c_client *client)
+{
+	struct tpm_chip *chip = (struct tpm_chip *)i2c_get_clientdata(client);
+	struct st33zp24_platform_data *pin_infos =
+		((struct i2c_client *) chip->vendor.iobase)->dev.platform_data;
+
+	if (pin_infos != NULL) {
+		free_irq(pin_infos->io_serirq, chip);
+
+		gpio_free(pin_infos->io_serirq);
+		gpio_free(pin_infos->io_lpcpd);
+
+		if (pin_infos->bChipF != true)
+			tpm_remove_hardware(chip->dev);
+		if (pin_infos->tpm_i2c_buffer[1] != NULL) {
+			kzfree(pin_infos->tpm_i2c_buffer[1]);
+			pin_infos->tpm_i2c_buffer[1] = NULL;
+		}
+		if (pin_infos->tpm_i2c_buffer[0] != NULL) {
+			kzfree(pin_infos->tpm_i2c_buffer[0]);
+			pin_infos->tpm_i2c_buffer[0] = NULL;
+		}
+	}
+
+	return 0;
+}
+
+/*
+ * tpm_st33_i2c_pm_suspend suspend the TPM device
+ * Added: Work around when suspend and no tpm application is running, suspend
+ * may fail because chip->data_buffer is not set (only set in tpm_open in Linux
+ * TPM core)
+ * @param: client, the i2c_client drescription (TPM I2C description).
+ * @param: mesg, the power management message.
+ * @return: 0 in case of success.
+ */
+static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
+{
+	struct tpm_chip *chip =
+	    (struct tpm_chip *)i2c_get_clientdata(client);
+	struct st33zp24_platform_data *pin_infos =
+		((struct i2c_client *)chip->vendor.iobase)->dev.platform_data;
+	int ret = 0;
+
+	if (power_mgt)
+		gpio_set_value(pin_infos->io_lpcpd, 0);
+	else{
+		if (chip->data_buffer == NULL)
+			chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
+		ret = tpm_pm_suspend(&client->dev, mesg);
+	}
+	return ret;
+}				/* tpm_st33_i2c_suspend() */
+
+/*
+ * tpm_st33_i2c_pm_resume resume the TPM device
+ * @param: client, the i2c_client drescription (TPM I2C description).
+ * @return: 0 in case of success.
+ */
+static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
+{
+	struct tpm_chip *chip =
+	    (struct tpm_chip *)i2c_get_clientdata(client);
+	struct st33zp24_platform_data *pin_infos =
+		((struct i2c_client *)chip->vendor.iobase)->dev.platform_data;
+
+	int ret = 0;
+
+	if (power_mgt) {
+		gpio_set_value(pin_infos->io_lpcpd, 1);
+		ret = wait_for_serirq_timeout(chip,
+					  (chip->vendor.status(chip) &&
+					  TPM_STS_VALID) == TPM_STS_VALID,
+					  chip->vendor.timeout_b);
+	} else{
+	if (chip->data_buffer == NULL)
+		chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
+	ret = tpm_pm_resume(&client->dev);
+	if (!ret)
+		tpm_continue_selftest(chip);
+	}
+	return ret;
+}				/* tpm_st33_i2c_pm_resume() */
+
+static const struct i2c_device_id tpm_st33_i2c_id[] = {
+	{TPM_ST33_I2C, 0},
+	{}
+};
+
+MODULE_DEVICE_TABLE(i2c, tpm_st33_i2c_id);
+
+static struct i2c_driver tpm_st33_i2c_driver = {
+	.driver = {
+		   .owner = THIS_MODULE,
+		   .name = TPM_ST33_I2C,
+		   },
+	.probe = tpm_st33_i2c_probe,
+	.remove = tpm_st33_i2c_remove,
+	.resume = tpm_st33_i2c_pm_resume,
+	.suspend = tpm_st33_i2c_pm_suspend,
+	.id_table = tpm_st33_i2c_id
+};
+
+/*
+ * tpm_st33_i2c_init initialize driver
+ * @return: 0 if successful, else non zero value.
+ */
+static int __init tpm_st33_i2c_init(void)
+{
+	return i2c_add_driver(&tpm_st33_i2c_driver);
+}
+
+/*
+ * tpm_st33_i2c_exit The kernel calls this function during unloading the
+ * module or during shut down process
+ */
+static void __exit tpm_st33_i2c_exit(void)
+{
+	i2c_del_driver(&tpm_st33_i2c_driver);
+}
+
+module_init(tpm_st33_i2c_init);
+module_exit(tpm_st33_i2c_exit);
+
+MODULE_AUTHOR("Christophe Ricard ([email protected])");
+MODULE_DESCRIPTION("STM TPM I2C ST33 Driver");
+MODULE_VERSION("1.2.0");
+MODULE_LICENSE("GPL");
drivers/char/tpm/tpm_stm_st33_i2c.h new file +81 -0

81 lines not on this page · raw message

include/linux/i2c/tpm_stm_st33_i2c.h new file +47 -0

47 lines not on this page · raw message

Signature · 1 line
1.7.9.5
#2 lore raw
In reply to Mathias Leblanc

Hi Mathias,

Quoted 17 lines · Mathias Leblanc

* STMicroelectronics version 1.2.0, Copyright (C) 2010
* STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
* This is free software, and you are welcome to redistribute it
* under certain conditions.

This is the driver for TPM chip from ST Microelectronics.

If you have a TPM security chip from STMicroelectronics working with an I2C, in menuconfig or .config choose the tpm driver on device --> tpm and activate the protocol of your choice before compiling the kernel.
The driver will be accessible from within Linux.

Tested on linux x86/x64, beagleboard REV B & XM REV C and CHROMIUM OS on kernel 3.x.x

Quoted 8 lines
---
 drivers/char/tpm/tpm_stm_st33_i2c.c  |  903 ++++++++++++++++++++++++++++++++++
 drivers/char/tpm/tpm_stm_st33_i2c.h  |   81 +++
 include/linux/i2c/tpm_stm_st33_i2c.h |   47 ++
 3 files changed, 1031 insertions(+)
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.c
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.h
 create mode 100644 include/linux/i2c/tpm_stm_st33_i2c.h

I'm still getting the compiler errors I mentioned earlier. I've fixed
them up with the patch below (plus some formatting), all except for the tpm_continue_selftest_nocheck() problem. Do you really need to call continue selftest on resume? That strikes me as out of place here. What's making the call to the TPM to start the self test in that case?

Please test with the patch below.

Thanks,
Kent

drivers/char/tpm/tpm_stm_st33_i2c.c +9 -9
index 52f3380..67f81b6 100644
@@ -722,7 +722,7 @@ tpm_st33_i2c_probe(struct i2c_client *client, const struct i2c_device_id *id)
 			goto _irq_set;
 
 		intmask = TPM_GLOBAL_INT_ENABLE;
-		err = I2C_WRITE_DATA(client, TPM_INT_ENABLE + 3, &intmask, 1);
+		err = I2C_WRITE_DATA(client, (TPM_INT_ENABLE + 3), &intmask, 1);
 		if (err < 0)
 			goto _irq_set;
 
@@ -820,10 +820,10 @@ static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
 
 	if (power_mgt)
 		gpio_set_value(pin_infos->io_lpcpd, 0);
-	else{
+	else {
 		if (chip->data_buffer == NULL)
 			chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-		ret = tpm_pm_suspend(&client->dev, mesg);
+		ret = tpm_pm_suspend(&client->dev);
 	}
 	return ret;
 }				/* tpm_st33_i2c_suspend() */
@@ -848,12 +848,12 @@ static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
 					  (chip->vendor.status(chip) &&
 					  TPM_STS_VALID) == TPM_STS_VALID,
 					  chip->vendor.timeout_b);
-	} else{
-	if (chip->data_buffer == NULL)
-		chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-	ret = tpm_pm_resume(&client->dev);
-	if (!ret)
-		tpm_continue_selftest_nocheck(chip);
+	} else {
+		if (chip->data_buffer == NULL)
+			chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
+		ret = tpm_pm_resume(&client->dev);
+		if (!ret)
+			tpm_continue_selftest_nocheck(chip);
 	}
 	return ret;
 }				/* tpm_st33_i2c_pm_resume() */
#3 lore raw

Hi Kent,

What's version of the kernel do you compiled?
I don't understand why you have some problem with formating, i have send you the patch with no error and warning of code formating. The tpm_continue_selftest_nocheck() (in kernel 3.x.x) and tpm_continue_selftest() (in kernel 2.6.x) is called to have the tpm ready. If the tpm_continue_selftest is not called by the driver, it will be use in the applications.

Regards,

Matthias Leblanc
________________________________________
From: Kent Yoder [[email protected]]
Sent: 26 November 2012 18:46
To: Mathias LEBLANC
Cc: Rajiv Andrade; Marcel Selhorst; Sirrix AG; [email protected]; Debora Velarde; [email protected]; Jean-Luc BLANC
Subject: Re: [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Hi Mathias,

Quoted 17 lines · Mathias Leblanc

* STMicroelectronics version 1.2.0, Copyright (C) 2010
* STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
* This is free software, and you are welcome to redistribute it
* under certain conditions.

This is the driver for TPM chip from ST Microelectronics.

If you have a TPM security chip from STMicroelectronics working with an I2C, in menuconfig or .config choose the tpm driver on device --> tpm and activate the protocol of your choice before compiling the kernel.
The driver will be accessible from within Linux.

Tested on linux x86/x64, beagleboard REV B & XM REV C and CHROMIUM OS on kernel 3.x.x

Quoted 8 lines
---
 drivers/char/tpm/tpm_stm_st33_i2c.c  |  903 ++++++++++++++++++++++++++++++++++
 drivers/char/tpm/tpm_stm_st33_i2c.h  |   81 +++
 include/linux/i2c/tpm_stm_st33_i2c.h |   47 ++
 3 files changed, 1031 insertions(+)
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.c
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.h
 create mode 100644 include/linux/i2c/tpm_stm_st33_i2c.h

I'm still getting the compiler errors I mentioned earlier. I've fixed
them up with the patch below (plus some formatting), all except for the tpm_continue_selftest_nocheck() problem. Do you really need to call continue selftest on resume? That strikes me as out of place here. What's making the call to the TPM to start the self test in that case?

Please test with the patch below.

Thanks,
Kent

drivers/char/tpm/tpm_stm_st33_i2c.c +9 -9
index 52f3380..67f81b6 100644
@@ -722,7 +722,7 @@ tpm_st33_i2c_probe(struct i2c_client *client, const struct i2c_device_id *id)
                        goto _irq_set;
 
                intmask = TPM_GLOBAL_INT_ENABLE;
-               err = I2C_WRITE_DATA(client, TPM_INT_ENABLE + 3, &intmask, 1);
+               err = I2C_WRITE_DATA(client, (TPM_INT_ENABLE + 3), &intmask, 1);
                if (err < 0)
                        goto _irq_set;
 
@@ -820,10 +820,10 @@ static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
 
        if (power_mgt)
                gpio_set_value(pin_infos->io_lpcpd, 0);
-       else{
+       else {
                if (chip->data_buffer == NULL)
                        chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-               ret = tpm_pm_suspend(&client->dev, mesg);
+               ret = tpm_pm_suspend(&client->dev);
        }
        return ret;
 }                              /* tpm_st33_i2c_suspend() */
@@ -848,12 +848,12 @@ static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
                                          (chip->vendor.status(chip) &&
                                          TPM_STS_VALID) == TPM_STS_VALID,
                                          chip->vendor.timeout_b);
-       } else{
-       if (chip->data_buffer == NULL)
-               chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-       ret = tpm_pm_resume(&client->dev);
-       if (!ret)
-               tpm_continue_selftest_nocheck(chip);
+       } else {
+               if (chip->data_buffer == NULL)
+                       chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
+               ret = tpm_pm_resume(&client->dev);
+               if (!ret)
+                       tpm_continue_selftest_nocheck(chip);
        }
        return ret;
 }                              /* tpm_st33_i2c_pm_resume() */
#4 lore raw
Quoted 6 lines · Mathias LEBLANC

Hi Kent,

What's version of the kernel do you compiled?
I don't understand why you have some problem with formating, i have send you the patch with no error and warning of code formating. The tpm_continue_selftest_nocheck() (in kernel 3.x.x) and tpm_continue_selftest() (in kernel 2.6.x) is called to have the tpm ready. If the tpm_continue_selftest is not called by the driver, it will be use in the applications.

Apply against this tree's "next" branch.

git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security.git

Thanks,
Kent

Quoted 107 lines

Regards,

Matthias Leblanc
________________________________________
From: Kent Yoder [[email protected]]
Sent: 26 November 2012 18:46
To: Mathias LEBLANC
Cc: Rajiv Andrade; Marcel Selhorst; Sirrix AG; [email protected]; Debora Velarde; [email protected]; Jean-Luc BLANC
Subject: Re: [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Hi Mathias,

Quoted 17 lines · Mathias Leblanc

* STMicroelectronics version 1.2.0, Copyright (C) 2010
* STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
* This is free software, and you are welcome to redistribute it
* under certain conditions.

This is the driver for TPM chip from ST Microelectronics.

If you have a TPM security chip from STMicroelectronics working with an I2C, in menuconfig or .config choose the tpm driver on device --> tpm and activate the protocol of your choice before compiling the kernel.
The driver will be accessible from within Linux.

Tested on linux x86/x64, beagleboard REV B & XM REV C and CHROMIUM OS on kernel 3.x.x

Quoted 8 lines
---
 drivers/char/tpm/tpm_stm_st33_i2c.c  |  903 ++++++++++++++++++++++++++++++++++
 drivers/char/tpm/tpm_stm_st33_i2c.h  |   81 +++
 include/linux/i2c/tpm_stm_st33_i2c.h |   47 ++
 3 files changed, 1031 insertions(+)
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.c
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.h
 create mode 100644 include/linux/i2c/tpm_stm_st33_i2c.h

I'm still getting the compiler errors I mentioned earlier. I've fixed
them up with the patch below (plus some formatting), all except for the tpm_continue_selftest_nocheck() problem. Do you really need to call continue selftest on resume? That strikes me as out of place here. What's making the call to the TPM to start the self test in that case?

Please test with the patch below.

Thanks,
Kent

drivers/char/tpm/tpm_stm_st33_i2c.c +9 -10
index 52f3380..67f81b6 100644
@@ -722,7 +722,7 @@ tpm_st33_i2c_probe(struct i2c_client *client, const struct i2c_device_id *id)
                        goto _irq_set;
 
                intmask = TPM_GLOBAL_INT_ENABLE;
-               err = I2C_WRITE_DATA(client, TPM_INT_ENABLE + 3, &intmask, 1);
+               err = I2C_WRITE_DATA(client, (TPM_INT_ENABLE + 3), &intmask, 1);
                if (err < 0)
                        goto _irq_set;
 
@@ -820,10 +820,10 @@ static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
 
        if (power_mgt)
                gpio_set_value(pin_infos->io_lpcpd, 0);
-       else{
+       else {
                if (chip->data_buffer == NULL)
                        chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-               ret = tpm_pm_suspend(&client->dev, mesg);
+               ret = tpm_pm_suspend(&client->dev);
        }
        return ret;
 }                              /* tpm_st33_i2c_suspend() */
@@ -848,12 +848,12 @@ static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
                                          (chip->vendor.status(chip) &&
                                          TPM_STS_VALID) == TPM_STS_VALID,
                                          chip->vendor.timeout_b);
-       } else{
-       if (chip->data_buffer == NULL)
-               chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-       ret = tpm_pm_resume(&client->dev);
-       if (!ret)
-               tpm_continue_selftest_nocheck(chip);
+       } else {
+               if (chip->data_buffer == NULL)
+                       chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
+               ret = tpm_pm_resume(&client->dev);
+               if (!ret)
+                       tpm_continue_selftest_nocheck(chip);
        }
        return ret;
 }                              /* tpm_st33_i2c_pm_resume() */
 
------------------------------------------------------------------------------

Monitor your physical, virtual and cloud infrastructure from a single web console. Get in-depth insight into apps, servers, databases, vmware, SAP, cloud infrastructure, etc. Download 30-day Free Trial. Pricing starts from $795 for 25 servers or applications!
http://p.sf.net/sfu/zoho_dev2dev_nov
_______________________________________________
tpmdd-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/tpmdd-devel

Signature · 1 line
IBM LTC Security
#5 lore raw

Hi Kent,

I have been tried the patch on your linux branch and indeed, i have an error about tpm_continue_seftest.
What is the kernel I should use to test officially my patch? Because i saw a lot of different version of linux kernel with the tpm_continue_seftest function. I used this repo git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6 for my tests.

Matthias Leblanc

________________________________________From: Kent Yoder [[email protected]]
Sent: 27 November 2012 15:48
To: Mathias LEBLANC
Cc: Kent Yoder; Jean-Luc BLANC; [email protected]; Rajiv Andrade; [email protected]; Sirrix AG
Subject: Re: [tpmdd-devel] [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Quoted 6 lines · Mathias LEBLANC

Hi Kent,

What's version of the kernel do you compiled?
I don't understand why you have some problem with formating, i have send you the patch with no error and warning of code formating. The tpm_continue_selftest_nocheck() (in kernel 3.x.x) and tpm_continue_selftest() (in kernel 2.6.x) is called to have the tpm ready. If the tpm_continue_selftest is not called by the driver, it will be use in the applications.

Apply against this tree's "next" branch.

git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security.git

Thanks,
Kent

Quoted 99 lines

Regards,

Matthias Leblanc
________________________________________
From: Kent Yoder [[email protected]]
Sent: 26 November 2012 18:46
To: Mathias LEBLANC
Cc: Rajiv Andrade; Marcel Selhorst; Sirrix AG; [email protected]; Debora Velarde; [email protected]; Jean-Luc BLANC
Subject: Re: [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Hi Mathias,

Quoted 17 lines · Mathias Leblanc

* STMicroelectronics version 1.2.0, Copyright (C) 2010
* STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
* This is free software, and you are welcome to redistribute it
* under certain conditions.

This is the driver for TPM chip from ST Microelectronics.

If you have a TPM security chip from STMicroelectronics working with an I2C, in menuconfig or .config choose the tpm driver on device --> tpm and activate the protocol of your choice before compiling the kernel.
The driver will be accessible from within Linux.

Tested on linux x86/x64, beagleboard REV B & XM REV C and CHROMIUM OS on kernel 3.x.x

Quoted 8 lines
---
 drivers/char/tpm/tpm_stm_st33_i2c.c  |  903 ++++++++++++++++++++++++++++++++++
 drivers/char/tpm/tpm_stm_st33_i2c.h  |   81 +++
 include/linux/i2c/tpm_stm_st33_i2c.h |   47 ++
 3 files changed, 1031 insertions(+)
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.c
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.h
 create mode 100644 include/linux/i2c/tpm_stm_st33_i2c.h

I'm still getting the compiler errors I mentioned earlier. I've fixed
them up with the patch below (plus some formatting), all except for the tpm_continue_selftest_nocheck() problem. Do you really need to call continue selftest on resume? That strikes me as out of place here. What's making the call to the TPM to start the self test in that case?

Please test with the patch below.

Thanks,
Kent

drivers/char/tpm/tpm_stm_st33_i2c.c +9 -10
index 52f3380..67f81b6 100644
@@ -722,7 +722,7 @@ tpm_st33_i2c_probe(struct i2c_client *client, const struct i2c_device_id *id)
                        goto _irq_set;
 
                intmask = TPM_GLOBAL_INT_ENABLE;
-               err = I2C_WRITE_DATA(client, TPM_INT_ENABLE + 3, &intmask, 1);
+               err = I2C_WRITE_DATA(client, (TPM_INT_ENABLE + 3), &intmask, 1);
                if (err < 0)
                        goto _irq_set;
 
@@ -820,10 +820,10 @@ static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
 
        if (power_mgt)
                gpio_set_value(pin_infos->io_lpcpd, 0);
-       else{
+       else {
                if (chip->data_buffer == NULL)
                        chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-               ret = tpm_pm_suspend(&client->dev, mesg);
+               ret = tpm_pm_suspend(&client->dev);
        }
        return ret;
 }                              /* tpm_st33_i2c_suspend() */
@@ -848,12 +848,12 @@ static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
                                          (chip->vendor.status(chip) &&
                                          TPM_STS_VALID) == TPM_STS_VALID,
                                          chip->vendor.timeout_b);
-       } else{
-       if (chip->data_buffer == NULL)
-               chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-       ret = tpm_pm_resume(&client->dev);
-       if (!ret)
-               tpm_continue_selftest_nocheck(chip);
+       } else {
+               if (chip->data_buffer == NULL)
+                       chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
+               ret = tpm_pm_resume(&client->dev);
+               if (!ret)
+                       tpm_continue_selftest_nocheck(chip);
        }
        return ret;
 }                              /* tpm_st33_i2c_pm_resume() */
 
------------------------------------------------------------------------------

Monitor your physical, virtual and cloud infrastructure from a single

Quoted 8 lines

web console. Get in-depth insight into apps, servers, databases, vmware, SAP, cloud infrastructure, etc. Download 30-day Free Trial. Pricing starts from $795 for 25 servers or applications!
http://p.sf.net/sfu/zoho_dev2dev_nov
_______________________________________________
tpmdd-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/tpmdd-devel

--
IBM LTC Security

#6 lore raw
Quoted 5 lines · Mathias LEBLANC

Hi Kent,

I have been tried the patch on your linux branch and indeed, i have an error about tpm_continue_seftest.
What is the kernel I should use to test officially my patch? Because i saw a lot of different version of linux kernel with the tpm_continue_seftest function. I used this repo git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6 for my tests.

The tree url I sent you is the correct one. Make sure you use the
"next" branch.

Kent

Quoted 135 lines

Matthias Leblanc

________________________________________From: Kent Yoder [[email protected]]
Sent: 27 November 2012 15:48
To: Mathias LEBLANC
Cc: Kent Yoder; Jean-Luc BLANC; [email protected]; Rajiv Andrade; [email protected]; Sirrix AG
Subject: Re: [tpmdd-devel] [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Quoted 6 lines · Mathias LEBLANC

Hi Kent,

What's version of the kernel do you compiled?
I don't understand why you have some problem with formating, i have send you the patch with no error and warning of code formating. The tpm_continue_selftest_nocheck() (in kernel 3.x.x) and tpm_continue_selftest() (in kernel 2.6.x) is called to have the tpm ready. If the tpm_continue_selftest is not called by the driver, it will be use in the applications.

Apply against this tree's "next" branch.

git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security.git

Thanks,
Kent

Quoted 99 lines

Regards,

Matthias Leblanc
________________________________________
From: Kent Yoder [[email protected]]
Sent: 26 November 2012 18:46
To: Mathias LEBLANC
Cc: Rajiv Andrade; Marcel Selhorst; Sirrix AG; [email protected]; Debora Velarde; [email protected]; Jean-Luc BLANC
Subject: Re: [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Hi Mathias,

Quoted 17 lines · Mathias Leblanc

* STMicroelectronics version 1.2.0, Copyright (C) 2010
* STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
* This is free software, and you are welcome to redistribute it
* under certain conditions.

This is the driver for TPM chip from ST Microelectronics.

If you have a TPM security chip from STMicroelectronics working with an I2C, in menuconfig or .config choose the tpm driver on device --> tpm and activate the protocol of your choice before compiling the kernel.
The driver will be accessible from within Linux.

Tested on linux x86/x64, beagleboard REV B & XM REV C and CHROMIUM OS on kernel 3.x.x

Quoted 8 lines
---
 drivers/char/tpm/tpm_stm_st33_i2c.c  |  903 ++++++++++++++++++++++++++++++++++
 drivers/char/tpm/tpm_stm_st33_i2c.h  |   81 +++
 include/linux/i2c/tpm_stm_st33_i2c.h |   47 ++
 3 files changed, 1031 insertions(+)
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.c
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.h
 create mode 100644 include/linux/i2c/tpm_stm_st33_i2c.h

I'm still getting the compiler errors I mentioned earlier. I've fixed
them up with the patch below (plus some formatting), all except for the tpm_continue_selftest_nocheck() problem. Do you really need to call continue selftest on resume? That strikes me as out of place here. What's making the call to the TPM to start the self test in that case?

Please test with the patch below.

Thanks,
Kent

drivers/char/tpm/tpm_stm_st33_i2c.c +9 -10
index 52f3380..67f81b6 100644
@@ -722,7 +722,7 @@ tpm_st33_i2c_probe(struct i2c_client *client, const struct i2c_device_id *id)
                        goto _irq_set;
 
                intmask = TPM_GLOBAL_INT_ENABLE;
-               err = I2C_WRITE_DATA(client, TPM_INT_ENABLE + 3, &intmask, 1);
+               err = I2C_WRITE_DATA(client, (TPM_INT_ENABLE + 3), &intmask, 1);
                if (err < 0)
                        goto _irq_set;
 
@@ -820,10 +820,10 @@ static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
 
        if (power_mgt)
                gpio_set_value(pin_infos->io_lpcpd, 0);
-       else{
+       else {
                if (chip->data_buffer == NULL)
                        chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-               ret = tpm_pm_suspend(&client->dev, mesg);
+               ret = tpm_pm_suspend(&client->dev);
        }
        return ret;
 }                              /* tpm_st33_i2c_suspend() */
@@ -848,12 +848,12 @@ static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
                                          (chip->vendor.status(chip) &&
                                          TPM_STS_VALID) == TPM_STS_VALID,
                                          chip->vendor.timeout_b);
-       } else{
-       if (chip->data_buffer == NULL)
-               chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-       ret = tpm_pm_resume(&client->dev);
-       if (!ret)
-               tpm_continue_selftest_nocheck(chip);
+       } else {
+               if (chip->data_buffer == NULL)
+                       chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
+               ret = tpm_pm_resume(&client->dev);
+               if (!ret)
+                       tpm_continue_selftest_nocheck(chip);
        }
        return ret;
 }                              /* tpm_st33_i2c_pm_resume() */
 
------------------------------------------------------------------------------

Monitor your physical, virtual and cloud infrastructure from a single

Quoted 8 lines

web console. Get in-depth insight into apps, servers, databases, vmware, SAP, cloud infrastructure, etc. Download 30-day Free Trial. Pricing starts from $795 for 25 servers or applications!
http://p.sf.net/sfu/zoho_dev2dev_nov
_______________________________________________
tpmdd-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/tpmdd-devel

--
IBM LTC Security

Signature · 1 line
IBM LTC Security
#7 lore raw

Ok, so i have patch the ST33 I2C driver on this branch and correct some errors.
I send you the patch for the kernel 3.x
I have no error on compilation, tell me if you have problems.
I have implemented the tpm_do_selftest function to get the tpm ready, but it can be removed
________________________________________
From: Kent Yoder [[email protected]]
Sent: 28 November 2012 16:31
To: Mathias LEBLANC
Cc: Kent Yoder; Jean-Luc BLANC; [email protected]; Rajiv Andrade; [email protected]; Sirrix AG
Subject: Re: [tpmdd-devel] [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Quoted 5 lines · Mathias LEBLANC

Hi Kent,

I have been tried the patch on your linux branch and indeed, i have an error about tpm_continue_seftest.
What is the kernel I should use to test officially my patch? Because i saw a lot of different version of linux kernel with the tpm_continue_seftest function. I used this repo git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6 for my tests.

The tree url I sent you is the correct one. Make sure you use the
"next" branch.

Kent

Quoted 120 lines

Matthias Leblanc

________________________________________From: Kent Yoder [[email protected]]
Sent: 27 November 2012 15:48
To: Mathias LEBLANC
Cc: Kent Yoder; Jean-Luc BLANC; [email protected]; Rajiv Andrade; [email protected]; Sirrix AG
Subject: Re: [tpmdd-devel] [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Quoted 6 lines · Mathias LEBLANC

Hi Kent,

What's version of the kernel do you compiled?
I don't understand why you have some problem with formating, i have send you the patch with no error and warning of code formating. The tpm_continue_selftest_nocheck() (in kernel 3.x.x) and tpm_continue_selftest() (in kernel 2.6.x) is called to have the tpm ready. If the tpm_continue_selftest is not called by the driver, it will be use in the applications.

Apply against this tree's "next" branch.

git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security.git

Thanks,
Kent

Quoted 97 lines

Regards,

Matthias Leblanc
________________________________________
From: Kent Yoder [[email protected]]
Sent: 26 November 2012 18:46
To: Mathias LEBLANC
Cc: Rajiv Andrade; Marcel Selhorst; Sirrix AG; [email protected]; Debora Velarde; [email protected]; Jean-Luc BLANC
Subject: Re: [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Hi Mathias,

Quoted 17 lines · Mathias Leblanc

* STMicroelectronics version 1.2.0, Copyright (C) 2010
* STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
* This is free software, and you are welcome to redistribute it
* under certain conditions.

This is the driver for TPM chip from ST Microelectronics.

If you have a TPM security chip from STMicroelectronics working with an I2C, in menuconfig or .config choose the tpm driver on device --> tpm and activate the protocol of your choice before compiling the kernel.
The driver will be accessible from within Linux.

Tested on linux x86/x64, beagleboard REV B & XM REV C and CHROMIUM OS on kernel 3.x.x

Quoted 8 lines
---
 drivers/char/tpm/tpm_stm_st33_i2c.c  |  903 ++++++++++++++++++++++++++++++++++
 drivers/char/tpm/tpm_stm_st33_i2c.h  |   81 +++
 include/linux/i2c/tpm_stm_st33_i2c.h |   47 ++
 3 files changed, 1031 insertions(+)
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.c
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.h
 create mode 100644 include/linux/i2c/tpm_stm_st33_i2c.h

I'm still getting the compiler errors I mentioned earlier. I've fixed
them up with the patch below (plus some formatting), all except for the tpm_continue_selftest_nocheck() problem. Do you really need to call continue selftest on resume? That strikes me as out of place here. What's making the call to the TPM to start the self test in that case?

Please test with the patch below.

Thanks,
Kent

drivers/char/tpm/tpm_stm_st33_i2c.c +9 -9
index 52f3380..67f81b6 100644
@@ -722,7 +722,7 @@ tpm_st33_i2c_probe(struct i2c_client *client, const struct i2c_device_id *id)
                        goto _irq_set;
 
                intmask = TPM_GLOBAL_INT_ENABLE;
-               err = I2C_WRITE_DATA(client, TPM_INT_ENABLE + 3, &intmask, 1);
+               err = I2C_WRITE_DATA(client, (TPM_INT_ENABLE + 3), &intmask, 1);
                if (err < 0)
                        goto _irq_set;
 
@@ -820,10 +820,10 @@ static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
 
        if (power_mgt)
                gpio_set_value(pin_infos->io_lpcpd, 0);
-       else{
+       else {
                if (chip->data_buffer == NULL)
                        chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-               ret = tpm_pm_suspend(&client->dev, mesg);
+               ret = tpm_pm_suspend(&client->dev);
        }
        return ret;
 }                              /* tpm_st33_i2c_suspend() */
@@ -848,12 +848,12 @@ static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
                                          (chip->vendor.status(chip) &&
                                          TPM_STS_VALID) == TPM_STS_VALID,
                                          chip->vendor.timeout_b);
-       } else{
-       if (chip->data_buffer == NULL)
-               chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-       ret = tpm_pm_resume(&client->dev);
-       if (!ret)
-               tpm_continue_selftest_nocheck(chip);
+       } else {
+               if (chip->data_buffer == NULL)
+                       chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
+               ret = tpm_pm_resume(&client->dev);
+               if (!ret)
+                       tpm_continue_selftest_nocheck(chip);
        }
        return ret;
 }                              /* tpm_st33_i2c_pm_resume() */
Quoted 15 lines
Quoted 10 lines

------------------------------------------------------------------------------
Monitor your physical, virtual and cloud infrastructure from a single web console. Get in-depth insight into apps, servers, databases, vmware, SAP, cloud infrastructure, etc. Download 30-day Free Trial. Pricing starts from $795 for 25 servers or applications!
http://p.sf.net/sfu/zoho_dev2dev_nov
_______________________________________________
tpmdd-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/tpmdd-devel

--
IBM LTC Security

--
IBM LTC Security

From 3cf4687665a0bcffe89de92416f0d5886a90100f Mon Sep 17 00:00:00 2001
From: Mathias Leblanc <[email protected]>
Date: Wed, 28 Nov 2012 18:22:24 +0100
Subject: [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x

* STMicroelectronics version 1.2.0, Copyright (C) 2010
* STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
* This is free software, and you are welcome to redistribute it
* under certain conditions.

This is the driver for TPM chip from ST Microelectronics.

If you have a TPM security chip from STMicroelectronics working with an I2C, in menuconfig or .config choose the tpm driver on device --> tpm and activate the protocol of your choice before compiling the kernel.
The driver will be accessible from within Linux.

Tested on linux x86/x64 on kernel 3.x

---
 drivers/char/tpm/tpm_stm_st33_i2c.c  |  903 ++++++++++++++++++++++++++++++++++
 drivers/char/tpm/tpm_stm_st33_i2c.h  |   81 +++
 include/linux/i2c/tpm_stm_st33_i2c.h |   47 ++
 3 files changed, 1031 insertions(+)
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.c
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.h
 create mode 100644 include/linux/i2c/tpm_stm_st33_i2c.h
drivers/char/tpm/tpm_stm_st33_i2c.c new file +903 -0
new file mode 100644
index 0000000..32cdd5e
@@ -0,0 +1,903 @@
+/*
+ * STMicroelectronics TPM I2C Linux driver for TPM ST33ZP24
+ * Copyright (C) 2009, 2010  STMicroelectronics
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * STMicroelectronics version 1.2.0, Copyright (C) 2010
+ * STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
+ * This is free software, and you are welcome to redistribute it
+ * under certain conditions.
+ *
+ * @Author: Christophe RICARD [email protected]
+ *
+ * @File: tpm_stm_st33_i2c.c
+ *
+ * @Synopsis:
+ *	09/15/2010:	First shot driver tpm_tis driver for
+			 lpc is used as model.
+ */
+
+#include <linux/i2c/tpm_stm_st33_i2c.h>
+
+#include "tpm_stm_st33_i2c.h"
+
+enum stm33zp24_access {
+	TPM_ACCESS_VALID = 0x80,
+	TPM_ACCESS_ACTIVE_LOCALITY = 0x20,
+	TPM_ACCESS_REQUEST_PENDING = 0x04,
+	TPM_ACCESS_REQUEST_USE = 0x02,
+};
+
+enum stm33zp24_status {
+	TPM_STS_VALID = 0x80,
+	TPM_STS_COMMAND_READY = 0x40,
+	TPM_STS_GO = 0x20,
+	TPM_STS_DATA_AVAIL = 0x10,
+	TPM_STS_DATA_EXPECT = 0x08,
+};
+
+enum stm33zp24_int_flags {
+	TPM_GLOBAL_INT_ENABLE = 0x80,
+	TPM_INTF_CMD_READY_INT = 0x080,
+	TPM_INTF_FIFO_AVALAIBLE_INT = 0x040,
+	TPM_INTF_WAKE_UP_READY_INT = 0x020,
+	TPM_INTF_LOCALITY_CHANGE_INT = 0x004,
+	TPM_INTF_STS_VALID_INT = 0x002,
+	TPM_INTF_DATA_AVAIL_INT = 0x001,
+};
+
+enum tis_defaults {
+	TIS_SHORT_TIMEOUT = 750,
+	TIS_LONG_TIMEOUT = 2000,
+};
+
+/*
+ * write8_reg
+ * Send byte to the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: tpm_register, the tpm tis register where the data should be written
+ * @param: tpm_data, the tpm_data to write inside the tpm_register
+ * @param: tpm_size, The length of the data
+ * @return: Returns negative errno, or else the number of bytes written.
+ */
+static int write8_reg(struct i2c_client *client, u8 tpm_register,
+		      u8 *tpm_data, u16 tpm_size)
+{
+	u8 data;
+	int value = 0;
+	struct st33zp24_platform_data *pin_infos;
+
+	pin_infos = client->dev.platform_data;
+
+	data = tpm_register;
+	memcpy(pin_infos->tpm_i2c_buffer[0], &data, sizeof(data));
+	memcpy(pin_infos->tpm_i2c_buffer[0] + 1, tpm_data, tpm_size);
+	value = i2c_master_send(client, pin_infos->tpm_i2c_buffer[0],
+				tpm_size + 1);
+	return value;
+} /* write8_reg() */
+
+/*
+ * read8_reg
+ * Recv byte from the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: tpm_register, the tpm tis register where the data should be read
+ * @param: tpm_data, the TPM response
+ * @param: tpm_size, tpm TPM response size to read.
+ * @return: number of byte read successfully: should be one if success.
+ */
+static int read8_reg(struct i2c_client *client, u8 tpm_register,
+		    u8 *tpm_data, int tpm_size)
+{
+	u8 status = 0;
+	u8 data;
+	struct st33zp24_platform_data *pin_infos;
+
+	pin_infos = client->dev.platform_data;
+
+	data = TPM_DUMMY_BYTE;
+	status = write8_reg(client, tpm_register, &data, 1);
+	if (status == 2)
+		status = i2c_master_recv(client, tpm_data, tpm_size);
+	return status;
+} /* read8_reg() */
+
+/*
+ * I2C_WRITE_DATA
+ * Send byte to the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: client, the chip description
+ * @param: tpm_register, the tpm tis register where the data should be written
+ * @param: tpm_data, the tpm_data to write inside the tpm_register
+ * @param: tpm_size, The length of the data
+ * @return: number of byte written successfully: should be one if success.
+ */
+#define I2C_WRITE_DATA(client, tpm_register, tpm_data, tpm_size) \
+	(write8_reg(client, tpm_register | \
+	TPM_WRITE_DIRECTION, tpm_data, tpm_size))
+
+/*
+ * I2C_READ_DATA
+ * Recv byte from the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: tpm, the chip description
+ * @param: tpm_register, the tpm tis register where the data should be read
+ * @param: tpm_data, the TPM response
+ * @param: tpm_size, tpm TPM response size to read.
+ * @return: number of byte read successfully: should be one if success.
+ */
+#define I2C_READ_DATA(client, tpm_register, tpm_data, tpm_size) \
+	(read8_reg(client, tpm_register, tpm_data, tpm_size))
+
+/*
+ * clear_interruption
+ * clear the TPM interrupt register.
+ * @param: tpm, the chip description
+ */
+static void clear_interruption(struct i2c_client *client)
+{
+	u8 interrupt;
+	I2C_READ_DATA(client, TPM_INT_STATUS, &interrupt, 1);
+	I2C_WRITE_DATA(client, TPM_INT_STATUS, &interrupt, 1);
+	I2C_READ_DATA(client, TPM_INT_STATUS, &interrupt, 1);
+} /* clear_interruption() */
+
+/*
+ * _wait_for_interrupt_serirq_timeout
+ * @param: tpm, the chip description
+ * @param: timeout, the timeout of the interrupt
+ * @return: the status of the interruption.
+ */
+static long _wait_for_interrupt_serirq_timeout(struct tpm_chip *chip,
+						unsigned long timeout)
+{
+	long status;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	status = wait_for_completion_interruptible_timeout(
+					&pin_infos->irq_detection,
+						timeout);
+	if (status > 0)
+		enable_irq(gpio_to_irq(pin_infos->io_serirq));
+	gpio_direction_input(pin_infos->io_serirq);
+
+	return status;
+} /* wait_for_interrupt_serirq_timeout() */
+
+int wait_for_serirq_timeout(struct tpm_chip *chip, bool condition,
+				 unsigned long timeout)
+{
+	int status = 2;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	status = _wait_for_interrupt_serirq_timeout(chip, timeout);
+	if (!status) {
+		status = -EBUSY;
+	} else{
+		clear_interruption(client);
+		if (condition)
+			status = 1;
+	}
+	return status;
+}
+
+/*
+ * tpm_stm_i2c_cancel, cancel is not implemented.
+ * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h
+ */
+static void tpm_stm_i2c_cancel(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	data = TPM_STS_COMMAND_READY;
+	I2C_WRITE_DATA(client, TPM_STS, &data, 1);
+	if (chip->vendor.irq)
+		wait_for_serirq_timeout(chip, 1, chip->vendor.timeout_a);
+}	/* tpm_stm_i2c_cancel() */
+
+/*
+ * tpm_stm_spi_status return the TPM_STS register
+ * @param: chip, the tpm chip description
+ * @return: the TPM_STS register value.
+ */
+static u8 tpm_stm_i2c_status(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	I2C_READ_DATA(client, TPM_STS, &data, 1);
+	return data;
+}				/* tpm_stm_i2c_status() */
+
+
+/*
+ * check_locality if the locality is active
+ * @param: chip, the tpm chip description
+ * @return: the active locality or -EACCESS.
+ */
+static int check_locality(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+	u8 status;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	status = I2C_READ_DATA(client, TPM_ACCESS, &data, 1);
+	if (status && (data &
+		(TPM_ACCESS_ACTIVE_LOCALITY | TPM_ACCESS_VALID)) ==
+		(TPM_ACCESS_ACTIVE_LOCALITY | TPM_ACCESS_VALID))
+		return chip->vendor.locality;
+
+	return -EACCES;
+
+} /* check_locality() */
+
+/*
+ * request_locality request the TPM locality
+ * @param: chip, the chip description
+ * @return: the active locality or EACCESS.
+ */
+static int request_locality(struct tpm_chip *chip)
+{
+	unsigned long stop;
+	long rc;
+	struct i2c_client *client;
+	u8 data;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	if (check_locality(chip) == chip->vendor.locality)
+		return chip->vendor.locality;
+
+	data = TPM_ACCESS_REQUEST_USE;
+	rc = I2C_WRITE_DATA(client, TPM_ACCESS, &data, 1);
+	if (rc < 0)
+		goto end;
+
+	if (chip->vendor.irq) {
+		rc = wait_for_serirq_timeout(chip, (check_locality
+						       (chip) >= 0),
+						      chip->vendor.timeout_a);
+		if (rc > 0)
+			return chip->vendor.locality;
+	} else{
+		stop = jiffies + chip->vendor.timeout_a;
+		do {
+			if (check_locality(chip) >= 0)
+				return chip->vendor.locality;
+			msleep(TPM_TIMEOUT);
+		} while (time_before(jiffies, stop));
+	}
+	rc = -EACCES;
+end:
+	return rc;
+} /* request_locality() */
+
+/*
+ * release_locality release the active locality
+ * @param: chip, the tpm chip description.
+ */
+static void release_locality(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	data = TPM_ACCESS_ACTIVE_LOCALITY;
+
+	I2C_WRITE_DATA(client, TPM_ACCESS, &data, 1);
+}
+
+/*
+ * get_burstcount return the burstcount address 0x19 0x1A
+ * @param: chip, the chip description
+ * return: the burstcount.
+ */
+static int get_burstcount(struct tpm_chip *chip)
+{
+	unsigned long stop;
+	int burstcnt, status;
+	u8 tpm_reg, temp;
+
+	struct i2c_client *client = (struct i2c_client *) chip->vendor.iobase;
+
+	stop = jiffies + chip->vendor.timeout_d;
+	do {
+		tpm_reg = TPM_STS + 1;
+		status = I2C_READ_DATA(client, tpm_reg, &temp, 1);
+		if (status < 0)
+			goto end;
+
+		tpm_reg = tpm_reg + 1;
+		burstcnt = temp;
+		status = I2C_READ_DATA(client, tpm_reg, &temp, 1);
+		if (status < 0)
+			goto end;
+
+		burstcnt |= temp << 8;
+		if (burstcnt)
+			return burstcnt;
+		msleep(TPM_TIMEOUT);
+	} while (time_before(jiffies, stop));
+
+end:
+	return -EBUSY;
+} /* get_burstcount() */
+
+/*
+ * wait_for_stat wait for a TPM_STS value
+ * @param: chip, the tpm chip description
+ * @param: mask, the value mask to wait
+ * @param: timeout, the timeout
+ * @param: queue, the wait queue.
+ * @return: the tpm status, 0 if success, -ETIME if timeout is reached.
+ */
+static int wait_for_stat(struct tpm_chip *chip, u8 mask, unsigned long timeout,
+			 wait_queue_head_t *queue)
+{
+	unsigned long stop;
+	long rc;
+	u8 status;
+
+	 if (chip->vendor.irq) {
+		rc = wait_for_serirq_timeout(chip, ((tpm_stm_i2c_status
+							(chip) & mask) ==
+						       mask), timeout);
+		if (rc > 0)
+			return 0;
+	} else{
+		stop = jiffies + timeout;
+		do {
+			msleep(TPM_TIMEOUT);
+			status = tpm_stm_i2c_status(chip);
+			if ((status & mask) == mask)
+				return 0;
+		} while (time_before(jiffies, stop));
+	}
+	return -ETIME;
+} /* wait_for_stat() */
+
+/*
+ * recv_data receive data
+ * @param: chip, the tpm chip description
+ * @param: buf, the buffer where the data are received
+ * @param: count, the number of data to receive
+ * @return: the number of bytes read from TPM FIFO.
+ */
+static int recv_data(struct tpm_chip *chip, u8 *buf, size_t count)
+{
+	int size = 0, burstcnt, len;
+	struct i2c_client *client;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	while (size < count &&
+	       wait_for_stat(chip,
+			     TPM_STS_DATA_AVAIL | TPM_STS_VALID,
+			     chip->vendor.timeout_c,
+			     &chip->vendor.read_queue)
+	       == 0) {
+		burstcnt = get_burstcount(chip);
+		len = min_t(int, burstcnt, count - size);
+		I2C_READ_DATA(client, TPM_DATA_FIFO, buf + size, len);
+		size += len;
+	}
+	return size;
+}
+
+/*
+ * tpm_ioserirq_handler the serirq irq handler
+ * @param: irq, the tpm chip description
+ * @param: dev_id, the description of the chip
+ * @return: the status of the handler.
+ */
+static irqreturn_t tpm_ioserirq_handler(int irq, void *dev_id)
+{
+	struct tpm_chip *chip = dev_id;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	disable_irq_nosync(irq);
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	complete(&pin_infos->irq_detection);
+	return IRQ_HANDLED;
+} /* tpm_ioserirq_handler() */
+
+
+/*
+ * tpm_stm_i2c_send send TPM commands through the I2C bus.
+ *
+ * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h
+ * @param: buf,	the buffer to send.
+ * @param: count, the number of bytes to send.
+ * @return: In case of success the number of bytes sent.
+ *			In other case, a < 0 value describing the issue.
+ */
+static int tpm_stm_i2c_send(struct tpm_chip *chip, unsigned char *buf,
+			    size_t len)
+{
+	u32 ret = 0, ordinal,
+	    status,
+	    burstcnt = 0, i, size;
+	u8 data;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	if (chip == NULL)
+		return -EBUSY;
+	if (len < TPM_HEADER_SIZE)
+		return -EBUSY;
+
+	client = (struct i2c_client *)chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	ordinal = be32_to_cpu(*((__be32 *) (buf + 6)));
+
+	client->flags = 0;
+
+	ret = request_locality(chip);
+	if (ret < 0)
+		return ret;
+
+	status = tpm_stm_i2c_status(chip);
+	if ((status & TPM_STS_COMMAND_READY) == 0) {
+		tpm_stm_i2c_cancel(chip);
+		if (wait_for_stat
+		    (chip, TPM_STS_COMMAND_READY, chip->vendor.timeout_b,
+		     &chip->vendor.int_queue) < 0) {
+			ret = -ETIME;
+			goto out_err;
+		}
+	}
+
+	for (i = 0 ; i < len - 1 ;) {
+		burstcnt = get_burstcount(chip);
+		size = min_t(int, len - i - 1, burstcnt);
+		ret = I2C_WRITE_DATA(client, TPM_DATA_FIFO, buf, size);
+		if (ret < 0)
+			goto out_err;
+
+		i += size;
+	}
+
+	status = tpm_stm_i2c_status(chip);
+	if ((status & TPM_STS_DATA_EXPECT) == 0) {
+		ret = -EIO;
+		goto out_err;
+	}
+
+	ret = I2C_WRITE_DATA(client, TPM_DATA_FIFO, buf + len - 1, 1);
+	if (ret < 0)
+		goto out_err;
+
+	status = tpm_stm_i2c_status(chip);
+	if ((status & TPM_STS_DATA_EXPECT) != 0) {
+		ret = -EIO;
+		goto out_err;
+	}
+
+	data = TPM_STS_GO;
+	I2C_WRITE_DATA(client, TPM_STS, &data, 1);
+
+	return len;
+out_err:
+	tpm_stm_i2c_cancel(chip);
+	release_locality(chip);
+	return ret;
+}
+
+/*
+ * tpm_stm_i2c_recv received TPM response through the I2C bus.
+ * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h.
+ * @param: buf,	the buffer to store datas.
+ * @param: count, the number of bytes to send.
+ * @return: In case of success the number of bytes received.
+ *		In other case, a < 0 value describing the issue.
+ */
+static int tpm_stm_i2c_recv(struct tpm_chip *chip, unsigned char *buf,
+			    size_t count)
+{
+	int size = 0;
+	int expected;
+
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	client = (struct i2c_client *)chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+
+	if (chip == NULL)
+		return -EBUSY;
+
+	if (count < TPM_HEADER_SIZE) {
+		size = -EIO;
+		goto out;
+	}
+
+	size = recv_data(chip, buf, TPM_HEADER_SIZE);
+	if (size < TPM_HEADER_SIZE) {
+		dev_err(chip->dev, "Unable to read header\n");
+		goto out;
+	}
+
+	expected = be32_to_cpu(*(__be32 *) (buf + 2));
+	if (expected > count) {
+		size = -EIO;
+		goto out;
+	}
+
+	size += recv_data(chip, &buf[TPM_HEADER_SIZE],
+					expected - TPM_HEADER_SIZE);
+	if (size < expected) {
+		dev_err(chip->dev, "Unable to read remainder of result\n");
+		size = -ETIME;
+		goto out;
+	}
+
+out:
+	chip->vendor.cancel(chip);
+	release_locality(chip);
+	return size;
+}
+
+static const struct file_operations tpm_st33_i2c_fops = {
+	.owner = THIS_MODULE,
+	.llseek = no_llseek,
+	.read = tpm_read,
+	.write = tpm_write,
+	.open = tpm_open,
+	.release = tpm_release,
+};
+
+static DEVICE_ATTR(pubek, S_IRUGO, tpm_show_pubek, NULL);
+static DEVICE_ATTR(pcrs, S_IRUGO, tpm_show_pcrs, NULL);
+static DEVICE_ATTR(enabled, S_IRUGO, tpm_show_enabled, NULL);
+static DEVICE_ATTR(active, S_IRUGO, tpm_show_active, NULL);
+static DEVICE_ATTR(owned, S_IRUGO, tpm_show_owned, NULL);
+static DEVICE_ATTR(temp_deactivated, S_IRUGO, tpm_show_temp_deactivated, NULL);
+static DEVICE_ATTR(caps, S_IRUGO, tpm_show_caps_1_2, NULL);
+static DEVICE_ATTR(cancel, S_IWUSR | S_IWGRP, NULL, tpm_store_cancel);
+
+static struct attribute *stm_tpm_attrs[] = {
+	&dev_attr_pubek.attr,
+	&dev_attr_pcrs.attr,
+	&dev_attr_enabled.attr,
+	&dev_attr_active.attr,
+	&dev_attr_owned.attr,
+	&dev_attr_temp_deactivated.attr,
+	&dev_attr_caps.attr,
+	&dev_attr_cancel.attr, NULL,
+};
+
+static struct attribute_group stm_tpm_attr_grp = {
+	.attrs = stm_tpm_attrs
+};
+
+static struct tpm_vendor_specific st_i2c_tpm = {
+	.send = tpm_stm_i2c_send,
+	.recv = tpm_stm_i2c_recv,
+	.cancel = tpm_stm_i2c_cancel,
+	.status = tpm_stm_i2c_status,
+	.req_complete_mask = TPM_STS_DATA_AVAIL | TPM_STS_VALID,
+	.req_complete_val = TPM_STS_DATA_AVAIL | TPM_STS_VALID,
+	.req_canceled = TPM_STS_COMMAND_READY,
+	.attr_group = &stm_tpm_attr_grp,
+	.miscdev = {.fops = &tpm_st33_i2c_fops,},
+};
+
+static int interrupts ;
+module_param(interrupts, int, 0444);
+MODULE_PARM_DESC(interrupts, "Enable interrupts");
+
+static int power_mgt = 1;
+module_param(power_mgt, int, 0444);
+MODULE_PARM_DESC(power_mgt, "Power Management");
+
+/*
+ * tpm_st33_i2c_probe initialize the TPM device
+ * @param: client, the i2c_client drescription (TPM I2C description).
+ * @param: id, the i2c_device_id struct.
+ * @return: 0 in case of success.
+ *	 -1 in other case.
+ */
+static int
+tpm_st33_i2c_probe(struct i2c_client *client, const struct i2c_device_id *id)
+{
+	u32 err;
+	u8 intmask;
+	struct tpm_chip *chip;
+	struct st33zp24_platform_data *platform_data;
+
+	err = 0;
+
+	if (client == NULL) {
+		dev_info(&client->dev, "client is NULL. exiting.\n");
+		err = -ENODEV;
+		goto end;
+	}
+
+	if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
+		dev_info(&client->dev, "client not i2c capable\n");
+		err = -ENODEV;
+		goto end;
+	}
+
+	chip = tpm_register_hardware(&client->dev, &st_i2c_tpm);
+	if (!chip) {
+		dev_info(&client->dev, "fail chip\n");
+		err = -ENODEV;
+		goto end;
+	}
+
+	platform_data = client->dev.platform_data;
+	platform_data->tpm_i2c_buffer[0] =
+	    kmalloc(TPM_BUFSIZE * sizeof(u8), GFP_KERNEL);
+	if (platform_data->tpm_i2c_buffer[0] == NULL) {
+		err = -ENOMEM;
+		goto _tpm_clean_answer;
+	}
+	platform_data->tpm_i2c_buffer[1] =
+	    kmalloc(TPM_BUFSIZE * sizeof(u8), GFP_KERNEL);
+	if (platform_data->tpm_i2c_buffer[1] == NULL) {
+		err = -ENOMEM;
+		goto _tpm_clean_response;
+	}
+
+	chip->vendor.iobase = client;
+
+	chip->vendor.timeout_a = msecs_to_jiffies(TIS_SHORT_TIMEOUT);
+	chip->vendor.timeout_b = msecs_to_jiffies(TIS_LONG_TIMEOUT);
+	chip->vendor.timeout_c = msecs_to_jiffies(TIS_SHORT_TIMEOUT);
+	chip->vendor.timeout_d = msecs_to_jiffies(TIS_SHORT_TIMEOUT);
+
+	chip->vendor.locality = LOCALITY0;
+
+	if (power_mgt) {
+		err = gpio_request(platform_data->io_lpcpd, "TPM IO_LPCPD");
+		if (err)
+			goto _gpio_init1;
+		gpio_set_value(platform_data->io_lpcpd, 1);
+	}
+
+	if (interrupts) {
+		init_completion(&platform_data->irq_detection);
+		if (request_locality(chip) != LOCALITY0) {
+			err = -ENODEV;
+			goto _tpm_clean_response;
+		}
+		err = gpio_request(platform_data->io_serirq, "TPM IO_SERIRQ");
+		if (err)
+			goto _gpio_init2;
+
+		clear_interruption(client);
+		err = request_irq(gpio_to_irq(platform_data->io_serirq),
+				&tpm_ioserirq_handler,
+				IRQF_TRIGGER_HIGH,
+				"TPM SERIRQ management", chip);
+		if (err < 0) {
+			dev_err(chip->dev , "TPM SERIRQ signals %d not available\n",
+			gpio_to_irq(platform_data->io_serirq));
+			goto _irq_set;
+		}
+
+		err = I2C_READ_DATA(client, TPM_INT_ENABLE, &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		intmask |= TPM_INTF_CMD_READY_INT
+			|  TPM_INTF_FIFO_AVALAIBLE_INT
+			|  TPM_INTF_WAKE_UP_READY_INT
+			|  TPM_INTF_LOCALITY_CHANGE_INT
+			|  TPM_INTF_STS_VALID_INT
+			|  TPM_INTF_DATA_AVAIL_INT;
+
+		err = I2C_WRITE_DATA(client, TPM_INT_ENABLE, &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		intmask = TPM_GLOBAL_INT_ENABLE;
+		err = I2C_WRITE_DATA(client, (TPM_INT_ENABLE + 3), &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		err = I2C_READ_DATA(client, TPM_INT_STATUS, &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		chip->vendor.irq = interrupts;
+
+		tpm_gen_interrupt(chip);
+	}
+
+	tpm_get_timeouts(chip);
+
+	i2c_set_clientdata(client, chip);
+	platform_data->bChipF = false;
+
+	dev_info(chip->dev, "TPM I2C Initialized\n");
+	return 0;
+_irq_set:
+	free_irq(gpio_to_irq(platform_data->io_serirq), (void *) chip);
+_gpio_init2:
+	if (platform_data && interrupts)
+		gpio_free(platform_data->io_serirq);
+_gpio_init1:
+	if (platform_data && power_mgt)
+		gpio_free(platform_data->io_lpcpd);
+_tpm_clean_response:
+	tpm_remove_hardware(chip->dev);
+	if (platform_data->tpm_i2c_buffer[1] != NULL) {
+		kzfree(platform_data->tpm_i2c_buffer[1]);
+		platform_data->tpm_i2c_buffer[1] = NULL;
+	}
+_tpm_clean_answer:
+	if (platform_data->tpm_i2c_buffer[0] != NULL) {
+		kzfree(platform_data->tpm_i2c_buffer[0]);
+		platform_data->tpm_i2c_buffer[0] = NULL;
+	}
+
+	platform_data->bChipF = true;
+end:
+	pr_info("TPM I2C initialisation fail\n");
+	return err;
+}
+
+/*
+ * tpm_st33_i2c_remove remove the TPM device
+ * @param: client, the i2c_client drescription (TPM I2C description).
+		clear_bit(0, &chip->is_open);
+ * @return: 0 in case of success.
+ */
+static __devexit int tpm_st33_i2c_remove(struct i2c_client *client)
+{
+	struct tpm_chip *chip = (struct tpm_chip *)i2c_get_clientdata(client);
+	struct st33zp24_platform_data *pin_infos =
+		((struct i2c_client *) chip->vendor.iobase)->dev.platform_data;
+
+	if (pin_infos != NULL) {
+		free_irq(pin_infos->io_serirq, chip);
+
+		gpio_free(pin_infos->io_serirq);
+		gpio_free(pin_infos->io_lpcpd);
+
+		if (pin_infos->bChipF != true)
+			tpm_remove_hardware(chip->dev);
+		if (pin_infos->tpm_i2c_buffer[1] != NULL) {
+			kzfree(pin_infos->tpm_i2c_buffer[1]);
+			pin_infos->tpm_i2c_buffer[1] = NULL;
+		}
+		if (pin_infos->tpm_i2c_buffer[0] != NULL) {
+			kzfree(pin_infos->tpm_i2c_buffer[0]);
+			pin_infos->tpm_i2c_buffer[0] = NULL;
+		}
+	}
+
+	return 0;
+}
+
+/*
+ * tpm_st33_i2c_pm_suspend suspend the TPM device
+ * Added: Work around when suspend and no tpm application is running, suspend
+ * may fail because chip->data_buffer is not set (only set in tpm_open in Linux
+ * TPM core)
+ * @param: client, the i2c_client drescription (TPM I2C description).
+ * @param: mesg, the power management message.
+ * @return: 0 in case of success.
+ */
+static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
+{
+	struct tpm_chip *chip =
+	    (struct tpm_chip *)i2c_get_clientdata(client);
+	struct st33zp24_platform_data *pin_infos =
+		((struct i2c_client *)chip->vendor.iobase)->dev.platform_data;
+	int ret = 0;
+
+	if (power_mgt)
+		gpio_set_value(pin_infos->io_lpcpd, 0);
+	else{
+		if (chip->data_buffer == NULL)
+			chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
+		ret = tpm_pm_suspend(&client->dev);
+	}
+	return ret;
+}				/* tpm_st33_i2c_suspend() */
+
+/*
+ * tpm_st33_i2c_pm_resume resume the TPM device
+ * @param: client, the i2c_client drescription (TPM I2C description).
+ * @return: 0 in case of success.
+ */
+static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
+{
+	struct tpm_chip *chip =
+	    (struct tpm_chip *)i2c_get_clientdata(client);
+	struct st33zp24_platform_data *pin_infos =
+		((struct i2c_client *)chip->vendor.iobase)->dev.platform_data;
+
+	int ret = 0;
+
+	if (power_mgt) {
+		gpio_set_value(pin_infos->io_lpcpd, 1);
+		ret = wait_for_serirq_timeout(chip,
+					  (chip->vendor.status(chip) &&
+					  TPM_STS_VALID) == TPM_STS_VALID,
+					  chip->vendor.timeout_b);
+	} else{
+	if (chip->data_buffer == NULL)
+		chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
+	ret = tpm_pm_resume(&client->dev);
+	if (!ret)
+		tpm_do_selftest(chip);
+	}
+	return ret;
+}				/* tpm_st33_i2c_pm_resume() */
+
+static const struct i2c_device_id tpm_st33_i2c_id[] = {
+	{TPM_ST33_I2C, 0},
+	{}
+};
+
+MODULE_DEVICE_TABLE(i2c, tpm_st33_i2c_id);
+
+static struct i2c_driver tpm_st33_i2c_driver = {
+	.driver = {
+		   .owner = THIS_MODULE,
+		   .name = TPM_ST33_I2C,
+		   },
+	.probe = tpm_st33_i2c_probe,
+	.remove = tpm_st33_i2c_remove,
+	.resume = tpm_st33_i2c_pm_resume,
+	.suspend = tpm_st33_i2c_pm_suspend,
+	.id_table = tpm_st33_i2c_id
+};
+
+/*
+ * tpm_st33_i2c_init initialize driver
+ * @return: 0 if successful, else non zero value.
+ */
+static int __init tpm_st33_i2c_init(void)
+{
+	return i2c_add_driver(&tpm_st33_i2c_driver);
+}
+
+/*
+ * tpm_st33_i2c_exit The kernel calls this function during unloading the
+ * module or during shut down process
+ */
+static void __exit tpm_st33_i2c_exit(void)
+{
+	i2c_del_driver(&tpm_st33_i2c_driver);
+}
+
+module_init(tpm_st33_i2c_init);
+module_exit(tpm_st33_i2c_exit);
+
+MODULE_AUTHOR("Christophe Ricard ([email protected])");
+MODULE_DESCRIPTION("STM TPM I2C ST33 Driver");
+MODULE_VERSION("1.2.0");
drivers/char/tpm/tpm_stm_st33_i2c.h new file +81 -0

81 lines not on this page · raw message

include/linux/i2c/tpm_stm_st33_i2c.h new file +47 -0

47 lines not on this page · raw message

Signature · 1 line
1.7.9.5
#8 lore raw
Quoted 4 lines · Mathias LEBLANC

Ok, so i have patch the ST33 I2C driver on this branch and correct some errors. I send you the patch for the kernel 3.x
I have no error on compilation, tell me if you have problems.
I have implemented the tpm_do_selftest function to get the tpm ready, but it can be removed

This looks like its base64 encoded...

Quoted 1,242 lines

________________________________________
From: Kent Yoder [[email protected]]
Sent: 28 November 2012 16:31
To: Mathias LEBLANC
Cc: Kent Yoder; Jean-Luc BLANC; [email protected]; Rajiv Andrade; [email protected]; Sirrix AG
Subject: Re: [tpmdd-devel] [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Quoted 5 lines · Mathias LEBLANC

Hi Kent,

I have been tried the patch on your linux branch and indeed, i have an error about tpm_continue_seftest.
What is the kernel I should use to test officially my patch? Because i saw a lot of different version of linux kernel with the tpm_continue_seftest function. I used this repo git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6 for my tests.

The tree url I sent you is the correct one. Make sure you use the
"next" branch.

Kent

Quoted 120 lines

Matthias Leblanc

________________________________________From: Kent Yoder [[email protected]]
Sent: 27 November 2012 15:48
To: Mathias LEBLANC
Cc: Kent Yoder; Jean-Luc BLANC; [email protected]; Rajiv Andrade; [email protected]; Sirrix AG
Subject: Re: [tpmdd-devel] [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Quoted 6 lines · Mathias LEBLANC

Hi Kent,

What's version of the kernel do you compiled?
I don't understand why you have some problem with formating, i have send you the patch with no error and warning of code formating. The tpm_continue_selftest_nocheck() (in kernel 3.x.x) and tpm_continue_selftest() (in kernel 2.6.x) is called to have the tpm ready. If the tpm_continue_selftest is not called by the driver, it will be use in the applications.

Apply against this tree's "next" branch.

git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security.git

Thanks,
Kent

Quoted 97 lines

Regards,

Matthias Leblanc
________________________________________
From: Kent Yoder [[email protected]]
Sent: 26 November 2012 18:46
To: Mathias LEBLANC
Cc: Rajiv Andrade; Marcel Selhorst; Sirrix AG; [email protected]; Debora Velarde; [email protected]; Jean-Luc BLANC
Subject: Re: [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Hi Mathias,

Quoted 17 lines · Mathias Leblanc

* STMicroelectronics version 1.2.0, Copyright (C) 2010
* STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
* This is free software, and you are welcome to redistribute it
* under certain conditions.

This is the driver for TPM chip from ST Microelectronics.

If you have a TPM security chip from STMicroelectronics working with an I2C, in menuconfig or .config choose the tpm driver on device --> tpm and activate the protocol of your choice before compiling the kernel.
The driver will be accessible from within Linux.

Tested on linux x86/x64, beagleboard REV B & XM REV C and CHROMIUM OS on kernel 3.x.x

Quoted 8 lines
---
 drivers/char/tpm/tpm_stm_st33_i2c.c  |  903 ++++++++++++++++++++++++++++++++++
 drivers/char/tpm/tpm_stm_st33_i2c.h  |   81 +++
 include/linux/i2c/tpm_stm_st33_i2c.h |   47 ++
 3 files changed, 1031 insertions(+)
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.c
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.h
 create mode 100644 include/linux/i2c/tpm_stm_st33_i2c.h

I'm still getting the compiler errors I mentioned earlier. I've fixed
them up with the patch below (plus some formatting), all except for the tpm_continue_selftest_nocheck() problem. Do you really need to call continue selftest on resume? That strikes me as out of place here. What's making the call to the TPM to start the self test in that case?

Please test with the patch below.

Thanks,
Kent

drivers/char/tpm/tpm_stm_st33_i2c.c +9 -9
index 52f3380..67f81b6 100644
@@ -722,7 +722,7 @@ tpm_st33_i2c_probe(struct i2c_client *client, const struct i2c_device_id *id)
                        goto _irq_set;
 
                intmask = TPM_GLOBAL_INT_ENABLE;
-               err = I2C_WRITE_DATA(client, TPM_INT_ENABLE + 3, &intmask, 1);
+               err = I2C_WRITE_DATA(client, (TPM_INT_ENABLE + 3), &intmask, 1);
                if (err < 0)
                        goto _irq_set;
 
@@ -820,10 +820,10 @@ static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
 
        if (power_mgt)
                gpio_set_value(pin_infos->io_lpcpd, 0);
-       else{
+       else {
                if (chip->data_buffer == NULL)
                        chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-               ret = tpm_pm_suspend(&client->dev, mesg);
+               ret = tpm_pm_suspend(&client->dev);
        }
        return ret;
 }                              /* tpm_st33_i2c_suspend() */
@@ -848,12 +848,12 @@ static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
                                          (chip->vendor.status(chip) &&
                                          TPM_STS_VALID) == TPM_STS_VALID,
                                          chip->vendor.timeout_b);
-       } else{
-       if (chip->data_buffer == NULL)
-               chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-       ret = tpm_pm_resume(&client->dev);
-       if (!ret)
-               tpm_continue_selftest_nocheck(chip);
+       } else {
+               if (chip->data_buffer == NULL)
+                       chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
+               ret = tpm_pm_resume(&client->dev);
+               if (!ret)
+                       tpm_continue_selftest_nocheck(chip);
        }
        return ret;
 }                              /* tpm_st33_i2c_pm_resume() */
Quoted 15 lines
Quoted 10 lines

------------------------------------------------------------------------------
Monitor your physical, virtual and cloud infrastructure from a single web console. Get in-depth insight into apps, servers, databases, vmware, SAP, cloud infrastructure, etc. Download 30-day Free Trial. Pricing starts from $795 for 25 servers or applications!
http://p.sf.net/sfu/zoho_dev2dev_nov
_______________________________________________
tpmdd-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/tpmdd-devel

--
IBM LTC Security

--
IBM LTC Security

From 3cf4687665a0bcffe89de92416f0d5886a90100f Mon Sep 17 00:00:00 2001
From: Mathias Leblanc <[email protected]>
Date: Wed, 28 Nov 2012 18:22:24 +0100
Subject: [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x

* STMicroelectronics version 1.2.0, Copyright (C) 2010
* STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
* This is free software, and you are welcome to redistribute it
* under certain conditions.

This is the driver for TPM chip from ST Microelectronics.

If you have a TPM security chip from STMicroelectronics working with an I2C, in menuconfig or .config choose the tpm driver on device --> tpm and activate the protocol of your choice before compiling the kernel.
The driver will be accessible from within Linux.

Tested on linux x86/x64 on kernel 3.x

---
 drivers/char/tpm/tpm_stm_st33_i2c.c  |  903 ++++++++++++++++++++++++++++++++++
 drivers/char/tpm/tpm_stm_st33_i2c.h  |   81 +++
 include/linux/i2c/tpm_stm_st33_i2c.h |   47 ++
 3 files changed, 1031 insertions(+)
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.c
 create mode 100644 drivers/char/tpm/tpm_stm_st33_i2c.h
 create mode 100644 include/linux/i2c/tpm_stm_st33_i2c.h
drivers/char/tpm/tpm_stm_st33_i2c.c new file +903 -0
new file mode 100644
index 0000000..32cdd5e
@@ -0,0 +1,903 @@
+/*
+ * STMicroelectronics TPM I2C Linux driver for TPM ST33ZP24
+ * Copyright (C) 2009, 2010  STMicroelectronics
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * STMicroelectronics version 1.2.0, Copyright (C) 2010
+ * STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
+ * This is free software, and you are welcome to redistribute it
+ * under certain conditions.
+ *
+ * @Author: Christophe RICARD [email protected]
+ *
+ * @File: tpm_stm_st33_i2c.c
+ *
+ * @Synopsis:
+ *	09/15/2010:	First shot driver tpm_tis driver for
+			 lpc is used as model.
+ */
+
+#include <linux/i2c/tpm_stm_st33_i2c.h>
+
+#include "tpm_stm_st33_i2c.h"
+
+enum stm33zp24_access {
+	TPM_ACCESS_VALID = 0x80,
+	TPM_ACCESS_ACTIVE_LOCALITY = 0x20,
+	TPM_ACCESS_REQUEST_PENDING = 0x04,
+	TPM_ACCESS_REQUEST_USE = 0x02,
+};
+
+enum stm33zp24_status {
+	TPM_STS_VALID = 0x80,
+	TPM_STS_COMMAND_READY = 0x40,
+	TPM_STS_GO = 0x20,
+	TPM_STS_DATA_AVAIL = 0x10,
+	TPM_STS_DATA_EXPECT = 0x08,
+};
+
+enum stm33zp24_int_flags {
+	TPM_GLOBAL_INT_ENABLE = 0x80,
+	TPM_INTF_CMD_READY_INT = 0x080,
+	TPM_INTF_FIFO_AVALAIBLE_INT = 0x040,
+	TPM_INTF_WAKE_UP_READY_INT = 0x020,
+	TPM_INTF_LOCALITY_CHANGE_INT = 0x004,
+	TPM_INTF_STS_VALID_INT = 0x002,
+	TPM_INTF_DATA_AVAIL_INT = 0x001,
+};
+
+enum tis_defaults {
+	TIS_SHORT_TIMEOUT = 750,
+	TIS_LONG_TIMEOUT = 2000,
+};
+
+/*
+ * write8_reg
+ * Send byte to the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: tpm_register, the tpm tis register where the data should be written
+ * @param: tpm_data, the tpm_data to write inside the tpm_register
+ * @param: tpm_size, The length of the data
+ * @return: Returns negative errno, or else the number of bytes written.
+ */
+static int write8_reg(struct i2c_client *client, u8 tpm_register,
+		      u8 *tpm_data, u16 tpm_size)
+{
+	u8 data;
+	int value = 0;
+	struct st33zp24_platform_data *pin_infos;
+
+	pin_infos = client->dev.platform_data;
+
+	data = tpm_register;
+	memcpy(pin_infos->tpm_i2c_buffer[0], &data, sizeof(data));
+	memcpy(pin_infos->tpm_i2c_buffer[0] + 1, tpm_data, tpm_size);
+	value = i2c_master_send(client, pin_infos->tpm_i2c_buffer[0],
+				tpm_size + 1);
+	return value;
+} /* write8_reg() */
+
+/*
+ * read8_reg
+ * Recv byte from the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: tpm_register, the tpm tis register where the data should be read
+ * @param: tpm_data, the TPM response
+ * @param: tpm_size, tpm TPM response size to read.
+ * @return: number of byte read successfully: should be one if success.
+ */
+static int read8_reg(struct i2c_client *client, u8 tpm_register,
+		    u8 *tpm_data, int tpm_size)
+{
+	u8 status = 0;
+	u8 data;
+	struct st33zp24_platform_data *pin_infos;
+
+	pin_infos = client->dev.platform_data;
+
+	data = TPM_DUMMY_BYTE;
+	status = write8_reg(client, tpm_register, &data, 1);
+	if (status == 2)
+		status = i2c_master_recv(client, tpm_data, tpm_size);
+	return status;
+} /* read8_reg() */
+
+/*
+ * I2C_WRITE_DATA
+ * Send byte to the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: client, the chip description
+ * @param: tpm_register, the tpm tis register where the data should be written
+ * @param: tpm_data, the tpm_data to write inside the tpm_register
+ * @param: tpm_size, The length of the data
+ * @return: number of byte written successfully: should be one if success.
+ */
+#define I2C_WRITE_DATA(client, tpm_register, tpm_data, tpm_size) \
+	(write8_reg(client, tpm_register | \
+	TPM_WRITE_DIRECTION, tpm_data, tpm_size))
+
+/*
+ * I2C_READ_DATA
+ * Recv byte from the TIS register according to the ST33ZP24 I2C protocol.
+ * @param: tpm, the chip description
+ * @param: tpm_register, the tpm tis register where the data should be read
+ * @param: tpm_data, the TPM response
+ * @param: tpm_size, tpm TPM response size to read.
+ * @return: number of byte read successfully: should be one if success.
+ */
+#define I2C_READ_DATA(client, tpm_register, tpm_data, tpm_size) \
+	(read8_reg(client, tpm_register, tpm_data, tpm_size))
+
+/*
+ * clear_interruption
+ * clear the TPM interrupt register.
+ * @param: tpm, the chip description
+ */
+static void clear_interruption(struct i2c_client *client)
+{
+	u8 interrupt;
+	I2C_READ_DATA(client, TPM_INT_STATUS, &interrupt, 1);
+	I2C_WRITE_DATA(client, TPM_INT_STATUS, &interrupt, 1);
+	I2C_READ_DATA(client, TPM_INT_STATUS, &interrupt, 1);
+} /* clear_interruption() */
+
+/*
+ * _wait_for_interrupt_serirq_timeout
+ * @param: tpm, the chip description
+ * @param: timeout, the timeout of the interrupt
+ * @return: the status of the interruption.
+ */
+static long _wait_for_interrupt_serirq_timeout(struct tpm_chip *chip,
+						unsigned long timeout)
+{
+	long status;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	status = wait_for_completion_interruptible_timeout(
+					&pin_infos->irq_detection,
+						timeout);
+	if (status > 0)
+		enable_irq(gpio_to_irq(pin_infos->io_serirq));
+	gpio_direction_input(pin_infos->io_serirq);
+
+	return status;
+} /* wait_for_interrupt_serirq_timeout() */
+
+int wait_for_serirq_timeout(struct tpm_chip *chip, bool condition,
+				 unsigned long timeout)
+{
+	int status = 2;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	status = _wait_for_interrupt_serirq_timeout(chip, timeout);
+	if (!status) {
+		status = -EBUSY;
+	} else{
+		clear_interruption(client);
+		if (condition)
+			status = 1;
+	}
+	return status;
+}
+
+/*
+ * tpm_stm_i2c_cancel, cancel is not implemented.
+ * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h
+ */
+static void tpm_stm_i2c_cancel(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	data = TPM_STS_COMMAND_READY;
+	I2C_WRITE_DATA(client, TPM_STS, &data, 1);
+	if (chip->vendor.irq)
+		wait_for_serirq_timeout(chip, 1, chip->vendor.timeout_a);
+}	/* tpm_stm_i2c_cancel() */
+
+/*
+ * tpm_stm_spi_status return the TPM_STS register
+ * @param: chip, the tpm chip description
+ * @return: the TPM_STS register value.
+ */
+static u8 tpm_stm_i2c_status(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	I2C_READ_DATA(client, TPM_STS, &data, 1);
+	return data;
+}				/* tpm_stm_i2c_status() */
+
+
+/*
+ * check_locality if the locality is active
+ * @param: chip, the tpm chip description
+ * @return: the active locality or -EACCESS.
+ */
+static int check_locality(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+	u8 status;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	status = I2C_READ_DATA(client, TPM_ACCESS, &data, 1);
+	if (status && (data &
+		(TPM_ACCESS_ACTIVE_LOCALITY | TPM_ACCESS_VALID)) ==
+		(TPM_ACCESS_ACTIVE_LOCALITY | TPM_ACCESS_VALID))
+		return chip->vendor.locality;
+
+	return -EACCES;
+
+} /* check_locality() */
+
+/*
+ * request_locality request the TPM locality
+ * @param: chip, the chip description
+ * @return: the active locality or EACCESS.
+ */
+static int request_locality(struct tpm_chip *chip)
+{
+	unsigned long stop;
+	long rc;
+	struct i2c_client *client;
+	u8 data;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	if (check_locality(chip) == chip->vendor.locality)
+		return chip->vendor.locality;
+
+	data = TPM_ACCESS_REQUEST_USE;
+	rc = I2C_WRITE_DATA(client, TPM_ACCESS, &data, 1);
+	if (rc < 0)
+		goto end;
+
+	if (chip->vendor.irq) {
+		rc = wait_for_serirq_timeout(chip, (check_locality
+						       (chip) >= 0),
+						      chip->vendor.timeout_a);
+		if (rc > 0)
+			return chip->vendor.locality;
+	} else{
+		stop = jiffies + chip->vendor.timeout_a;
+		do {
+			if (check_locality(chip) >= 0)
+				return chip->vendor.locality;
+			msleep(TPM_TIMEOUT);
+		} while (time_before(jiffies, stop));
+	}
+	rc = -EACCES;
+end:
+	return rc;
+} /* request_locality() */
+
+/*
+ * release_locality release the active locality
+ * @param: chip, the tpm chip description.
+ */
+static void release_locality(struct tpm_chip *chip)
+{
+	struct i2c_client *client;
+	u8 data;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	data = TPM_ACCESS_ACTIVE_LOCALITY;
+
+	I2C_WRITE_DATA(client, TPM_ACCESS, &data, 1);
+}
+
+/*
+ * get_burstcount return the burstcount address 0x19 0x1A
+ * @param: chip, the chip description
+ * return: the burstcount.
+ */
+static int get_burstcount(struct tpm_chip *chip)
+{
+	unsigned long stop;
+	int burstcnt, status;
+	u8 tpm_reg, temp;
+
+	struct i2c_client *client = (struct i2c_client *) chip->vendor.iobase;
+
+	stop = jiffies + chip->vendor.timeout_d;
+	do {
+		tpm_reg = TPM_STS + 1;
+		status = I2C_READ_DATA(client, tpm_reg, &temp, 1);
+		if (status < 0)
+			goto end;
+
+		tpm_reg = tpm_reg + 1;
+		burstcnt = temp;
+		status = I2C_READ_DATA(client, tpm_reg, &temp, 1);
+		if (status < 0)
+			goto end;
+
+		burstcnt |= temp << 8;
+		if (burstcnt)
+			return burstcnt;
+		msleep(TPM_TIMEOUT);
+	} while (time_before(jiffies, stop));
+
+end:
+	return -EBUSY;
+} /* get_burstcount() */
+
+/*
+ * wait_for_stat wait for a TPM_STS value
+ * @param: chip, the tpm chip description
+ * @param: mask, the value mask to wait
+ * @param: timeout, the timeout
+ * @param: queue, the wait queue.
+ * @return: the tpm status, 0 if success, -ETIME if timeout is reached.
+ */
+static int wait_for_stat(struct tpm_chip *chip, u8 mask, unsigned long timeout,
+			 wait_queue_head_t *queue)
+{
+	unsigned long stop;
+	long rc;
+	u8 status;
+
+	 if (chip->vendor.irq) {
+		rc = wait_for_serirq_timeout(chip, ((tpm_stm_i2c_status
+							(chip) & mask) ==
+						       mask), timeout);
+		if (rc > 0)
+			return 0;
+	} else{
+		stop = jiffies + timeout;
+		do {
+			msleep(TPM_TIMEOUT);
+			status = tpm_stm_i2c_status(chip);
+			if ((status & mask) == mask)
+				return 0;
+		} while (time_before(jiffies, stop));
+	}
+	return -ETIME;
+} /* wait_for_stat() */
+
+/*
+ * recv_data receive data
+ * @param: chip, the tpm chip description
+ * @param: buf, the buffer where the data are received
+ * @param: count, the number of data to receive
+ * @return: the number of bytes read from TPM FIFO.
+ */
+static int recv_data(struct tpm_chip *chip, u8 *buf, size_t count)
+{
+	int size = 0, burstcnt, len;
+	struct i2c_client *client;
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+
+	while (size < count &&
+	       wait_for_stat(chip,
+			     TPM_STS_DATA_AVAIL | TPM_STS_VALID,
+			     chip->vendor.timeout_c,
+			     &chip->vendor.read_queue)
+	       == 0) {
+		burstcnt = get_burstcount(chip);
+		len = min_t(int, burstcnt, count - size);
+		I2C_READ_DATA(client, TPM_DATA_FIFO, buf + size, len);
+		size += len;
+	}
+	return size;
+}
+
+/*
+ * tpm_ioserirq_handler the serirq irq handler
+ * @param: irq, the tpm chip description
+ * @param: dev_id, the description of the chip
+ * @return: the status of the handler.
+ */
+static irqreturn_t tpm_ioserirq_handler(int irq, void *dev_id)
+{
+	struct tpm_chip *chip = dev_id;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	disable_irq_nosync(irq);
+
+	client = (struct i2c_client *) chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	complete(&pin_infos->irq_detection);
+	return IRQ_HANDLED;
+} /* tpm_ioserirq_handler() */
+
+
+/*
+ * tpm_stm_i2c_send send TPM commands through the I2C bus.
+ *
+ * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h
+ * @param: buf,	the buffer to send.
+ * @param: count, the number of bytes to send.
+ * @return: In case of success the number of bytes sent.
+ *			In other case, a < 0 value describing the issue.
+ */
+static int tpm_stm_i2c_send(struct tpm_chip *chip, unsigned char *buf,
+			    size_t len)
+{
+	u32 ret = 0, ordinal,
+	    status,
+	    burstcnt = 0, i, size;
+	u8 data;
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	if (chip == NULL)
+		return -EBUSY;
+	if (len < TPM_HEADER_SIZE)
+		return -EBUSY;
+
+	client = (struct i2c_client *)chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+	ordinal = be32_to_cpu(*((__be32 *) (buf + 6)));
+
+	client->flags = 0;
+
+	ret = request_locality(chip);
+	if (ret < 0)
+		return ret;
+
+	status = tpm_stm_i2c_status(chip);
+	if ((status & TPM_STS_COMMAND_READY) == 0) {
+		tpm_stm_i2c_cancel(chip);
+		if (wait_for_stat
+		    (chip, TPM_STS_COMMAND_READY, chip->vendor.timeout_b,
+		     &chip->vendor.int_queue) < 0) {
+			ret = -ETIME;
+			goto out_err;
+		}
+	}
+
+	for (i = 0 ; i < len - 1 ;) {
+		burstcnt = get_burstcount(chip);
+		size = min_t(int, len - i - 1, burstcnt);
+		ret = I2C_WRITE_DATA(client, TPM_DATA_FIFO, buf, size);
+		if (ret < 0)
+			goto out_err;
+
+		i += size;
+	}
+
+	status = tpm_stm_i2c_status(chip);
+	if ((status & TPM_STS_DATA_EXPECT) == 0) {
+		ret = -EIO;
+		goto out_err;
+	}
+
+	ret = I2C_WRITE_DATA(client, TPM_DATA_FIFO, buf + len - 1, 1);
+	if (ret < 0)
+		goto out_err;
+
+	status = tpm_stm_i2c_status(chip);
+	if ((status & TPM_STS_DATA_EXPECT) != 0) {
+		ret = -EIO;
+		goto out_err;
+	}
+
+	data = TPM_STS_GO;
+	I2C_WRITE_DATA(client, TPM_STS, &data, 1);
+
+	return len;
+out_err:
+	tpm_stm_i2c_cancel(chip);
+	release_locality(chip);
+	return ret;
+}
+
+/*
+ * tpm_stm_i2c_recv received TPM response through the I2C bus.
+ * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h.
+ * @param: buf,	the buffer to store datas.
+ * @param: count, the number of bytes to send.
+ * @return: In case of success the number of bytes received.
+ *		In other case, a < 0 value describing the issue.
+ */
+static int tpm_stm_i2c_recv(struct tpm_chip *chip, unsigned char *buf,
+			    size_t count)
+{
+	int size = 0;
+	int expected;
+
+	struct i2c_client *client;
+	struct st33zp24_platform_data *pin_infos;
+
+	client = (struct i2c_client *)chip->vendor.iobase;
+	pin_infos = client->dev.platform_data;
+
+
+	if (chip == NULL)
+		return -EBUSY;
+
+	if (count < TPM_HEADER_SIZE) {
+		size = -EIO;
+		goto out;
+	}
+
+	size = recv_data(chip, buf, TPM_HEADER_SIZE);
+	if (size < TPM_HEADER_SIZE) {
+		dev_err(chip->dev, "Unable to read header\n");
+		goto out;
+	}
+
+	expected = be32_to_cpu(*(__be32 *) (buf + 2));
+	if (expected > count) {
+		size = -EIO;
+		goto out;
+	}
+
+	size += recv_data(chip, &buf[TPM_HEADER_SIZE],
+					expected - TPM_HEADER_SIZE);
+	if (size < expected) {
+		dev_err(chip->dev, "Unable to read remainder of result\n");
+		size = -ETIME;
+		goto out;
+	}
+
+out:
+	chip->vendor.cancel(chip);
+	release_locality(chip);
+	return size;
+}
+
+static const struct file_operations tpm_st33_i2c_fops = {
+	.owner = THIS_MODULE,
+	.llseek = no_llseek,
+	.read = tpm_read,
+	.write = tpm_write,
+	.open = tpm_open,
+	.release = tpm_release,
+};
+
+static DEVICE_ATTR(pubek, S_IRUGO, tpm_show_pubek, NULL);
+static DEVICE_ATTR(pcrs, S_IRUGO, tpm_show_pcrs, NULL);
+static DEVICE_ATTR(enabled, S_IRUGO, tpm_show_enabled, NULL);
+static DEVICE_ATTR(active, S_IRUGO, tpm_show_active, NULL);
+static DEVICE_ATTR(owned, S_IRUGO, tpm_show_owned, NULL);
+static DEVICE_ATTR(temp_deactivated, S_IRUGO, tpm_show_temp_deactivated, NULL);
+static DEVICE_ATTR(caps, S_IRUGO, tpm_show_caps_1_2, NULL);
+static DEVICE_ATTR(cancel, S_IWUSR | S_IWGRP, NULL, tpm_store_cancel);
+
+static struct attribute *stm_tpm_attrs[] = {
+	&dev_attr_pubek.attr,
+	&dev_attr_pcrs.attr,
+	&dev_attr_enabled.attr,
+	&dev_attr_active.attr,
+	&dev_attr_owned.attr,
+	&dev_attr_temp_deactivated.attr,
+	&dev_attr_caps.attr,
+	&dev_attr_cancel.attr, NULL,
+};
+
+static struct attribute_group stm_tpm_attr_grp = {
+	.attrs = stm_tpm_attrs
+};
+
+static struct tpm_vendor_specific st_i2c_tpm = {
+	.send = tpm_stm_i2c_send,
+	.recv = tpm_stm_i2c_recv,
+	.cancel = tpm_stm_i2c_cancel,
+	.status = tpm_stm_i2c_status,
+	.req_complete_mask = TPM_STS_DATA_AVAIL | TPM_STS_VALID,
+	.req_complete_val = TPM_STS_DATA_AVAIL | TPM_STS_VALID,
+	.req_canceled = TPM_STS_COMMAND_READY,
+	.attr_group = &stm_tpm_attr_grp,
+	.miscdev = {.fops = &tpm_st33_i2c_fops,},
+};
+
+static int interrupts ;
+module_param(interrupts, int, 0444);
+MODULE_PARM_DESC(interrupts, "Enable interrupts");
+
+static int power_mgt = 1;
+module_param(power_mgt, int, 0444);
+MODULE_PARM_DESC(power_mgt, "Power Management");
+
+/*
+ * tpm_st33_i2c_probe initialize the TPM device
+ * @param: client, the i2c_client drescription (TPM I2C description).
+ * @param: id, the i2c_device_id struct.
+ * @return: 0 in case of success.
+ *	 -1 in other case.
+ */
+static int
+tpm_st33_i2c_probe(struct i2c_client *client, const struct i2c_device_id *id)
+{
+	u32 err;
+	u8 intmask;
+	struct tpm_chip *chip;
+	struct st33zp24_platform_data *platform_data;
+
+	err = 0;
+
+	if (client == NULL) {
+		dev_info(&client->dev, "client is NULL. exiting.\n");
+		err = -ENODEV;
+		goto end;
+	}
+
+	if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
+		dev_info(&client->dev, "client not i2c capable\n");
+		err = -ENODEV;
+		goto end;
+	}
+
+	chip = tpm_register_hardware(&client->dev, &st_i2c_tpm);
+	if (!chip) {
+		dev_info(&client->dev, "fail chip\n");
+		err = -ENODEV;
+		goto end;
+	}
+
+	platform_data = client->dev.platform_data;
+	platform_data->tpm_i2c_buffer[0] =
+	    kmalloc(TPM_BUFSIZE * sizeof(u8), GFP_KERNEL);
+	if (platform_data->tpm_i2c_buffer[0] == NULL) {
+		err = -ENOMEM;
+		goto _tpm_clean_answer;
+	}
+	platform_data->tpm_i2c_buffer[1] =
+	    kmalloc(TPM_BUFSIZE * sizeof(u8), GFP_KERNEL);
+	if (platform_data->tpm_i2c_buffer[1] == NULL) {
+		err = -ENOMEM;
+		goto _tpm_clean_response;
+	}
+
+	chip->vendor.iobase = client;
+
+	chip->vendor.timeout_a = msecs_to_jiffies(TIS_SHORT_TIMEOUT);
+	chip->vendor.timeout_b = msecs_to_jiffies(TIS_LONG_TIMEOUT);
+	chip->vendor.timeout_c = msecs_to_jiffies(TIS_SHORT_TIMEOUT);
+	chip->vendor.timeout_d = msecs_to_jiffies(TIS_SHORT_TIMEOUT);
+
+	chip->vendor.locality = LOCALITY0;
+
+	if (power_mgt) {
+		err = gpio_request(platform_data->io_lpcpd, "TPM IO_LPCPD");
+		if (err)
+			goto _gpio_init1;
+		gpio_set_value(platform_data->io_lpcpd, 1);
+	}
+
+	if (interrupts) {
+		init_completion(&platform_data->irq_detection);
+		if (request_locality(chip) != LOCALITY0) {
+			err = -ENODEV;
+			goto _tpm_clean_response;
+		}
+		err = gpio_request(platform_data->io_serirq, "TPM IO_SERIRQ");
+		if (err)
+			goto _gpio_init2;
+
+		clear_interruption(client);
+		err = request_irq(gpio_to_irq(platform_data->io_serirq),
+				&tpm_ioserirq_handler,
+				IRQF_TRIGGER_HIGH,
+				"TPM SERIRQ management", chip);
+		if (err < 0) {
+			dev_err(chip->dev , "TPM SERIRQ signals %d not available\n",
+			gpio_to_irq(platform_data->io_serirq));
+			goto _irq_set;
+		}
+
+		err = I2C_READ_DATA(client, TPM_INT_ENABLE, &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		intmask |= TPM_INTF_CMD_READY_INT
+			|  TPM_INTF_FIFO_AVALAIBLE_INT
+			|  TPM_INTF_WAKE_UP_READY_INT
+			|  TPM_INTF_LOCALITY_CHANGE_INT
+			|  TPM_INTF_STS_VALID_INT
+			|  TPM_INTF_DATA_AVAIL_INT;
+
+		err = I2C_WRITE_DATA(client, TPM_INT_ENABLE, &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		intmask = TPM_GLOBAL_INT_ENABLE;
+		err = I2C_WRITE_DATA(client, (TPM_INT_ENABLE + 3), &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		err = I2C_READ_DATA(client, TPM_INT_STATUS, &intmask, 1);
+		if (err < 0)
+			goto _irq_set;
+
+		chip->vendor.irq = interrupts;
+
+		tpm_gen_interrupt(chip);
+	}
+
+	tpm_get_timeouts(chip);
+
+	i2c_set_clientdata(client, chip);
+	platform_data->bChipF = false;
+
+	dev_info(chip->dev, "TPM I2C Initialized\n");
+	return 0;
+_irq_set:
+	free_irq(gpio_to_irq(platform_data->io_serirq), (void *) chip);
+_gpio_init2:
+	if (platform_data && interrupts)
+		gpio_free(platform_data->io_serirq);
+_gpio_init1:
+	if (platform_data && power_mgt)
+		gpio_free(platform_data->io_lpcpd);
+_tpm_clean_response:
+	tpm_remove_hardware(chip->dev);
+	if (platform_data->tpm_i2c_buffer[1] != NULL) {
+		kzfree(platform_data->tpm_i2c_buffer[1]);
+		platform_data->tpm_i2c_buffer[1] = NULL;
+	}
+_tpm_clean_answer:
+	if (platform_data->tpm_i2c_buffer[0] != NULL) {
+		kzfree(platform_data->tpm_i2c_buffer[0]);
+		platform_data->tpm_i2c_buffer[0] = NULL;
+	}
+
+	platform_data->bChipF = true;
+end:
+	pr_info("TPM I2C initialisation fail\n");
+	return err;
+}
+
+/*
+ * tpm_st33_i2c_remove remove the TPM device
+ * @param: client, the i2c_client drescription (TPM I2C description).
+		clear_bit(0, &chip->is_open);
+ * @return: 0 in case of success.
+ */
+static __devexit int tpm_st33_i2c_remove(struct i2c_client *client)
+{
+	struct tpm_chip *chip = (struct tpm_chip *)i2c_get_clientdata(client);
+	struct st33zp24_platform_data *pin_infos =
+		((struct i2c_client *) chip->vendor.iobase)->dev.platform_data;
+
+	if (pin_infos != NULL) {
+		free_irq(pin_infos->io_serirq, chip);
+
+		gpio_free(pin_infos->io_serirq);
+		gpio_free(pin_infos->io_lpcpd);
+
+		if (pin_infos->bChipF != true)
+			tpm_remove_hardware(chip->dev);
+		if (pin_infos->tpm_i2c_buffer[1] != NULL) {
+			kzfree(pin_infos->tpm_i2c_buffer[1]);
+			pin_infos->tpm_i2c_buffer[1] = NULL;
+		}
+		if (pin_infos->tpm_i2c_buffer[0] != NULL) {
+			kzfree(pin_infos->tpm_i2c_buffer[0]);
+			pin_infos->tpm_i2c_buffer[0] = NULL;
+		}
+	}
+
+	return 0;
+}
+
+/*
+ * tpm_st33_i2c_pm_suspend suspend the TPM device
+ * Added: Work around when suspend and no tpm application is running, suspend
+ * may fail because chip->data_buffer is not set (only set in tpm_open in Linux
+ * TPM core)
+ * @param: client, the i2c_client drescription (TPM I2C description).
+ * @param: mesg, the power management message.
+ * @return: 0 in case of success.
+ */
+static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
+{
+	struct tpm_chip *chip =
+	    (struct tpm_chip *)i2c_get_clientdata(client);
+	struct st33zp24_platform_data *pin_infos =
+		((struct i2c_client *)chip->vendor.iobase)->dev.platform_data;
+	int ret = 0;
+
+	if (power_mgt)
+		gpio_set_value(pin_infos->io_lpcpd, 0);
+	else{
+		if (chip->data_buffer == NULL)
+			chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
+		ret = tpm_pm_suspend(&client->dev);
+	}
+	return ret;
+}				/* tpm_st33_i2c_suspend() */
+
+/*
+ * tpm_st33_i2c_pm_resume resume the TPM device
+ * @param: client, the i2c_client drescription (TPM I2C description).
+ * @return: 0 in case of success.
+ */
+static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
+{
+	struct tpm_chip *chip =
+	    (struct tpm_chip *)i2c_get_clientdata(client);
+	struct st33zp24_platform_data *pin_infos =
+		((struct i2c_client *)chip->vendor.iobase)->dev.platform_data;
+
+	int ret = 0;
+
+	if (power_mgt) {
+		gpio_set_value(pin_infos->io_lpcpd, 1);
+		ret = wait_for_serirq_timeout(chip,
+					  (chip->vendor.status(chip) &&
+					  TPM_STS_VALID) == TPM_STS_VALID,
+					  chip->vendor.timeout_b);
+	} else{
+	if (chip->data_buffer == NULL)
+		chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
+	ret = tpm_pm_resume(&client->dev);
+	if (!ret)
+		tpm_do_selftest(chip);
+	}
+	return ret;
+}				/* tpm_st33_i2c_pm_resume() */
+
+static const struct i2c_device_id tpm_st33_i2c_id[] = {
+	{TPM_ST33_I2C, 0},
+	{}
+};
+
+MODULE_DEVICE_TABLE(i2c, tpm_st33_i2c_id);
+
+static struct i2c_driver tpm_st33_i2c_driver = {
+	.driver = {
+		   .owner = THIS_MODULE,
+		   .name = TPM_ST33_I2C,
+		   },
+	.probe = tpm_st33_i2c_probe,
+	.remove = tpm_st33_i2c_remove,
+	.resume = tpm_st33_i2c_pm_resume,
+	.suspend = tpm_st33_i2c_pm_suspend,
+	.id_table = tpm_st33_i2c_id
+};
+
+/*
+ * tpm_st33_i2c_init initialize driver
+ * @return: 0 if successful, else non zero value.
+ */
+static int __init tpm_st33_i2c_init(void)
+{
+	return i2c_add_driver(&tpm_st33_i2c_driver);
+}
+
+/*
+ * tpm_st33_i2c_exit The kernel calls this function during unloading the
+ * module or during shut down process
+ */
+static void __exit tpm_st33_i2c_exit(void)
+{
+	i2c_del_driver(&tpm_st33_i2c_driver);
+}
+
+module_init(tpm_st33_i2c_init);
+module_exit(tpm_st33_i2c_exit);
+
+MODULE_AUTHOR("Christophe Ricard ([email protected])");
+MODULE_DESCRIPTION("STM TPM I2C ST33 Driver");
+MODULE_VERSION("1.2.0");
drivers/char/tpm/tpm_stm_st33_i2c.h new file +81 -0
new file mode 100644
index 0000000..d390542
@@ -0,0 +1,81 @@
+/*
+ * STMicroelectronics TPM I2C Linux driver for TPM ST33ZP24
+ * Copyright (C) 2009, 2010  STMicroelectronics
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * STMicroelectronics version 1.2.0, Copyright (C) 2010
+ * STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
+ * This is free software, and you are welcome to redistribute it
+ * under certain conditions.
+ *
+ * @Author: Christophe RICARD [email protected]
+ *
+ * @File: stm_st33_tpm_i2c.h
+ *
+ * @Date: 09/15/2010
+ */
+#ifndef __STM_ST33_TPM_I2C_MAIN_H__
+#define __STM_ST33_TPM_I2C_MAIN_H__
+
+#include <linux/pci.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/i2c.h>
+#include <linux/fs.h>
+#include <linux/miscdevice.h>
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/delay.h>
+#include <linux/init.h>
+#include <linux/wait.h>
+#include <linux/string.h>
+#include <linux/interrupt.h>
+#include <linux/spinlock.h>
+#include <linux/sysfs.h>
+#include <linux/gpio.h>
+#include <linux/sched.h>
+#include <linux/uaccess.h>
+#include <linux/io.h>
+#include <linux/slab.h>
+#include <linux/sched.h>
+
+#include "tpm.h"
+
+#define MINOR_NUM_I2C		224
+
+#define TPM_ACCESS			(0x0)
+#define TPM_STS				(0x18)
+#define TPM_HASH_END			(0x20)
+#define TPM_DATA_FIFO			(0x24)
+#define TPM_HASH_DATA			(0x24)
+#define TPM_HASH_START			(0x28)
+#define TPM_INTF_CAPABILITY		(0x14)
+#define TPM_INT_STATUS			(0x10)
+#define TPM_INT_ENABLE			(0x08)
+
+#define TPM_DUMMY_BYTE			0xAA
+#define TPM_WRITE_DIRECTION		0x80
+#define TPM_HEADER_SIZE			10
+#define TPM_BUFSIZE			2048
+
+#define LOCALITY0		0
+
+struct st_tpm_hash {
+	int size;
+	u8 *data;
+};
+
+#endif /* __STM_ST33_TPM_I2C_MAIN_H__ */
include/linux/i2c/tpm_stm_st33_i2c.h new file +47 -0
new file mode 100644
index 0000000..e3ce42d
@@ -0,0 +1,47 @@
+/*
+* STMicroelectronics TPM I2C Linux driver for TPM ST33ZP24
+* Copyright (C) 2009, 2010 STMicroelectronics
+* Christophe RICARD [email protected]
+* This program is free software; you can redistribute it and/or modify
+* it under the terms of the GNU General Public License as published by
+* the Free Software Foundation; either version 2 of the License, or
+* (at your option) any later version.
+*
+* This program is distributed in the hope that it will be useful,
+* but WITHOUT ANY WARRANTY; without even the implied warranty of
+* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+* GNU General Public License for more details.
+*
+* You should have received a copy of the GNU General Public License along
+* with this program; if not, write to the Free Software Foundation, Inc.,
+* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+*
+* STMicroelectronics version 1.2.0, Copyright (C) 2010
+* STMicroelectronics comes with ABSOLUTELY NO WARRANTY.
+* This is free software, and you are welcome to redistribute it
+* under certain conditions.
+*
+* @File: stm_st33_tpm_i2c.h
+*
+* @Date: 06/15/2008
+*/
+#ifndef __STM_ST33_TPM_I2C_H__
+#define __STM_ST33_TPM_I2C_H__
+
+#include <linux/i2c.h>
+
+#define TPM_ST33_I2C			"st33zp24_i2c"
+#define TPM_I2C_ST33_ADDR_WR_LOC0	(0x26 >> 1)
+
+
+struct st33zp24_platform_data {
+	int io_serirq;
+	int io_lpcpd;
+	struct i2c_client *client;
+	bool bChipF;
+	u8 *tpm_i2c_buffer[2]; /* 0 Request 1 Response */
+	struct completion irq_detection;
+	struct mutex lock;
+};
+
+#endif /* __STM_ST33_TPM_I2C_H__ */
Signature · 1 line
1.7.9.5
maintainer
#9 lore raw
In reply to Mathias LEBLANC

Hi Mathias,

please note:
I'm writing this email on behalf of myself only and nobody else, especially not my employer - and I'm doing this in my spare time. I'm working for a direct competitor of yours, but I'm not using any knowledge that I've picked up at work or that is considered secret in any way. I have a personal interest in the TPM subsystem and want to keep it as clean as possible.
So please don't see my review as something negative, but rather something positive.

Am Mittwoch, 28. November 2012, 18:48:57 schrieb Mathias LEBLANC:

Quoted 5 lines

Ok, so i have patch the ST33 I2C driver on this branch and correct some errors. I send you the patch for the kernel 3.x
I have no error on compilation, tell me if you have problems. I have implemented the tpm_do_selftest function to get the tpm ready, but it can be removed ________________________________________

Unfortunately you attached the patch instead of sending it in plaintext which is the usual practice -> care to resend in plain text? Makes the review far easier.

(btw.: Please also have a look at
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmittingPatches;hb=HEAD
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmitChecklist;hb=HEAD
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmittingDrivers;hb=HEAD
which describes the process in detail)

When you resend the patch, can you please include the "metadata" as well - i.e. your modifications to the Kconfig / Makefile etc. I do not see a reason why to keep it in a seperate patch.

I tried the patch you've posted and it applies cleanly and now (finally) compiles as well - so now I can start with my review:

= The name =
There's already one i2c tpm driver in the tree, so maybe it would be a good idea to keep the naming scheme consistent? -> How about tpm_i2c_stm_st33.c ?
Eventually this is something Kent as a maintainer has to decice - but I would really like to see the name change. I hope we can eventually consolidate all the 'tis' based drivers.

= Compiling / License =
When compiling the driver I get the following warning
WARNING: modpost: missing MODULE_LICENSE() in /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o
Please include the appropriate MODULE_LICENSE as my kernel otherwise gets tainted by your driver.

Also this:
+ * STMicroelectronics TPM I2C Linux driver for TPM ST33ZP24
+ * Copyright (C) 2009, 2010 STMicroelectronics
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.

is not possible afaik - kernel code must be under GPL v2 _only_ without the "or (at your option) any later version." addition.

= sparse warnings =
When running sparse against your code I get the following warnings:
make -C /data/data-old/linux-2.6/ M=pwd modules C=1
make: Entering directory `/data/data-old/linux-2.6'
CHECK /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:167:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:187:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:180:5: warning: symbol 'wait_for_serirq_timeout' was not declared. Should it be static? /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:210:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:227:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:245:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:269:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:307:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:324:38: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:394:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:424:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:456:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:531:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:672:29: warning: incorrect type in assignment (different address spaces)
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:672:29: expected void [noderef] <asn:2>*iobase
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:672:29: got struct i2c_client *client
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:781:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:818:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:841:19: warning: cast removes address space of expression
CC [M] /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o
Building modules, stage 2.
MODPOST 8 modules

Please fix these if applicable - otherwise you'll probably get a friendly reminder to do so by fengguang's build test ;)

= smatch =
Same applies to smatch
make -C /data/data-old/linux-2.6/ M=pwd modules C=1 CHECK=smatch
make: Entering directory `/data/data-old/linux-2.6'
CHECK /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:535 tpm_stm_i2c_recv() warn: variable dereferenced before check 'chip' (see line 531) /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:748 tpm_st33_i2c_probe() warn: variable dereferenced before check 'platform_data' (see line 659) /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:848 tpm_st33_i2c_pm_resume() warn: should this be a bitwise op? /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:848 tpm_st33_i2c_pm_resume() warn: should this be a bitwise op?
CC [M] /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o

Please fix these if applicable - otherwise you'll probably get a friendly reminder to do so by fengguang's build test ;)

= checkpatch =
Also please run .../scripts/checkpatch.pl -strict before submission - not everything that is reported might be applicable, but quite often it is.

Looking forward to your v2 so I can give a more detailed code review of your code.

Thanks,
Peter

#10 lore raw

Hi Peter,

Thanks for your contribution.
I have modified the driver files name and descriptions. Regarding the warnings, it's strange.

@Kent, could you confirm that you have the same?

Regards,

Mathias Leblanc

-----Original Message-----
From: Peter Hüwe [mailto:[email protected]]
Sent: 29 November, 2012 01:05
To: Mathias LEBLANC
Cc: Kent Yoder; Kent Yoder; Jean-Luc BLANC; [email protected]; Rajiv Andrade; [email protected]; [email protected]
Subject: Re: [tpmdd-devel] [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Hi Mathias,

please note:
I'm writing this email on behalf of myself only and nobody else, especially not my employer - and I'm doing this in my spare time. I'm working for a direct competitor of yours, but I'm not using any knowledge that I've picked up at work or that is considered secret in any way. I have a personal interest in the TPM subsystem and want to keep it as clean as possible.
So please don't see my review as something negative, but rather something positive.

Am Mittwoch, 28. November 2012, 18:48:57 schrieb Mathias LEBLANC:

Quoted 5 lines

Ok, so i have patch the ST33 I2C driver on this branch and correct some errors. I send you the patch for the kernel 3.x I have no error on compilation, tell me if you have problems.
I have implemented the tpm_do_selftest function to get the tpm ready, but it can be removed ________________________________________

Unfortunately you attached the patch instead of sending it in plaintext which is the usual practice -> care to resend in plain text? Makes the review far easier.

(btw.: Please also have a look at
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmittingPatches;hb=HEAD
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmitChecklist;hb=HEAD
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmittingDrivers;hb=HEAD
which describes the process in detail)

When you resend the patch, can you please include the "metadata" as well - i.e. your modifications to the Kconfig / Makefile etc. I do not see a reason why to keep it in a seperate patch.

I tried the patch you've posted and it applies cleanly and now (finally) compiles as well - so now I can start with my review:

= The name =
There's already one i2c tpm driver in the tree, so maybe it would be a good idea to keep the naming scheme consistent? -> How about tpm_i2c_stm_st33.c ?
Eventually this is something Kent as a maintainer has to decice - but I would really like to see the name change. I hope we can eventually consolidate all the 'tis' based drivers.

= Compiling / License =
When compiling the driver I get the following warning
WARNING: modpost: missing MODULE_LICENSE() in /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o
Please include the appropriate MODULE_LICENSE as my kernel otherwise gets tainted by your driver.

Also this:
+ * STMicroelectronics TPM I2C Linux driver for TPM ST33ZP24
+ * Copyright (C) 2009, 2010 STMicroelectronics
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.

is not possible afaik - kernel code must be under GPL v2 _only_ without the "or (at your option) any later version." addition.

= sparse warnings =
When running sparse against your code I get the following warnings:
make -C /data/data-old/linux-2.6/ M=pwd modules C=1
make: Entering directory `/data/data-old/linux-2.6'
CHECK /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:167:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:187:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:180:5: warning: symbol 'wait_for_serirq_timeout' was not declared. Should it be static? /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:210:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:227:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:245:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:269:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:307:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:324:38: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:394:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:424:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:456:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:531:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:672:29: warning: incorrect type in assignment (different address spaces)
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:672:29: expected void [noderef] <asn:2>*iobase
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:672:29: got struct i2c_client *client
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:781:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:818:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:841:19: warning: cast removes address space of expression
CC [M] /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o
Building modules, stage 2.
MODPOST 8 modules

Please fix these if applicable - otherwise you'll probably get a friendly reminder to do so by fengguang's build test ;)

= smatch =
Same applies to smatch
make -C /data/data-old/linux-2.6/ M=pwd modules C=1 CHECK=smatch
make: Entering directory `/data/data-old/linux-2.6'
CHECK /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:535 tpm_stm_i2c_recv() warn: variable dereferenced before check 'chip' (see line 531) /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:748 tpm_st33_i2c_probe() warn: variable dereferenced before check 'platform_data' (see line 659) /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:848 tpm_st33_i2c_pm_resume() warn: should this be a bitwise op? /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:848 tpm_st33_i2c_pm_resume() warn: should this be a bitwise op?
CC [M] /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o

Please fix these if applicable - otherwise you'll probably get a friendly reminder to do so by fengguang's build test ;)

= checkpatch =
Also please run .../scripts/checkpatch.pl -strict before submission - not everything that is reported might be applicable, but quite often it is.

Looking forward to your v2 so I can give a more detailed code review of your code.

Thanks,
Peter

#11 lore raw
Quoted 7 lines · Mathias LEBLANC

Hi Peter,

Thanks for your contribution.
I have modified the driver files name and descriptions. Regarding the warnings, it's strange.

@Kent, could you confirm that you have the same?

Yep, I see these too. They're a side effect of storing the i2c_client
pointer in vendor->iobase, which is __iomem. The i2c_client struct isn't __iomem at all though, writes in the i2c subsystem are done through client->addr. So i2c_client should really be stored somewhere else, no?

vendor->data seems like the right place. I'm on the hook to update
the other drivers with a macro for accessing vendor->data in tpm.h. I'll make these updates and send out a patch.

Kent

Quoted 126 lines

Regards,

Mathias Leblanc

-----Original Message-----
From: Peter Hüwe [mailto:[email protected]]
Sent: 29 November, 2012 01:05
To: Mathias LEBLANC
Cc: Kent Yoder; Kent Yoder; Jean-Luc BLANC; [email protected]; Rajiv Andrade; [email protected]; [email protected]
Subject: Re: [tpmdd-devel] [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Hi Mathias,

please note:
I'm writing this email on behalf of myself only and nobody else, especially not my employer - and I'm doing this in my spare time. I'm working for a direct competitor of yours, but I'm not using any knowledge that I've picked up at work or that is considered secret in any way. I have a personal interest in the TPM subsystem and want to keep it as clean as possible.
So please don't see my review as something negative, but rather something positive.

Am Mittwoch, 28. November 2012, 18:48:57 schrieb Mathias LEBLANC:

Quoted 5 lines

Ok, so i have patch the ST33 I2C driver on this branch and correct some errors. I send you the patch for the kernel 3.x I have no error on compilation, tell me if you have problems.
I have implemented the tpm_do_selftest function to get the tpm ready, but it can be removed ________________________________________

Unfortunately you attached the patch instead of sending it in plaintext which is the usual practice -> care to resend in plain text? Makes the review far easier.

(btw.: Please also have a look at
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmittingPatches;hb=HEAD
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmitChecklist;hb=HEAD
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmittingDrivers;hb=HEAD
which describes the process in detail)

When you resend the patch, can you please include the "metadata" as well - i.e. your modifications to the Kconfig / Makefile etc. I do not see a reason why to keep it in a seperate patch.

I tried the patch you've posted and it applies cleanly and now (finally) compiles as well - so now I can start with my review:

= The name =
There's already one i2c tpm driver in the tree, so maybe it would be a good idea to keep the naming scheme consistent? -> How about tpm_i2c_stm_st33.c ?
Eventually this is something Kent as a maintainer has to decice - but I would really like to see the name change. I hope we can eventually consolidate all the 'tis' based drivers.

= Compiling / License =
When compiling the driver I get the following warning
WARNING: modpost: missing MODULE_LICENSE() in /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o
Please include the appropriate MODULE_LICENSE as my kernel otherwise gets tainted by your driver.

Also this:
+ * STMicroelectronics TPM I2C Linux driver for TPM ST33ZP24
+ * Copyright (C) 2009, 2010 STMicroelectronics
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.

is not possible afaik - kernel code must be under GPL v2 _only_ without the "or (at your option) any later version." addition.

= sparse warnings =
When running sparse against your code I get the following warnings:
make -C /data/data-old/linux-2.6/ M=pwd modules C=1
make: Entering directory `/data/data-old/linux-2.6'
CHECK /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:167:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:187:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:180:5: warning: symbol 'wait_for_serirq_timeout' was not declared. Should it be static? /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:210:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:227:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:245:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:269:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:307:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:324:38: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:394:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:424:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:456:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:531:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:672:29: warning: incorrect type in assignment (different address spaces)
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:672:29: expected void [noderef] <asn:2>*iobase
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:672:29: got struct i2c_client *client
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:781:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:818:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:841:19: warning: cast removes address space of expression
CC [M] /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o
Building modules, stage 2.
MODPOST 8 modules

Please fix these if applicable - otherwise you'll probably get a friendly reminder to do so by fengguang's build test ;)

= smatch =
Same applies to smatch
make -C /data/data-old/linux-2.6/ M=pwd modules C=1 CHECK=smatch
make: Entering directory `/data/data-old/linux-2.6'
CHECK /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:535 tpm_stm_i2c_recv() warn: variable dereferenced before check 'chip' (see line 531) /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:748 tpm_st33_i2c_probe() warn: variable dereferenced before check 'platform_data' (see line 659) /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:848 tpm_st33_i2c_pm_resume() warn: should this be a bitwise op? /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:848 tpm_st33_i2c_pm_resume() warn: should this be a bitwise op?
CC [M] /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o

Please fix these if applicable - otherwise you'll probably get a friendly reminder to do so by fengguang's build test ;)

= checkpatch =
Also please run .../scripts/checkpatch.pl -strict before submission - not everything that is reported might be applicable, but quite often it is.

Looking forward to your v2 so I can give a more detailed code review of your code.

Thanks,
Peter

#12 lore raw
Quoted 15 lines · Kent Yoder
Quoted 7 lines · Mathias LEBLANC

Hi Peter,

Thanks for your contribution.
I have modified the driver files name and descriptions. Regarding the warnings, it's strange.

@Kent, could you confirm that you have the same?

Yep, I see these too. They're a side effect of storing the i2c_client
pointer in vendor->iobase, which is __iomem. The i2c_client struct isn't __iomem at all though, writes in the i2c subsystem are done through client->addr. So i2c_client should really be stored somewhere else, no?

Quoted 3 lines

vendor->data seems like the right place. I'm on the hook to update
the other drivers with a macro for accessing vendor->data in tpm.h. I'll make these updates and send out a patch.

Mathias, one other thing, MODULE_LICENSE(): GPL?

Kent

Quoted 141 lines

Kent

Quoted 119 lines

Regards,

Mathias Leblanc

-----Original Message-----
From: Peter Hüwe [mailto:[email protected]]
Sent: 29 November, 2012 01:05
To: Mathias LEBLANC
Cc: Kent Yoder; Kent Yoder; Jean-Luc BLANC; [email protected]; Rajiv Andrade; [email protected]; [email protected]
Subject: Re: [tpmdd-devel] [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Hi Mathias,

please note:
I'm writing this email on behalf of myself only and nobody else, especially not my employer - and I'm doing this in my spare time. I'm working for a direct competitor of yours, but I'm not using any knowledge that I've picked up at work or that is considered secret in any way. I have a personal interest in the TPM subsystem and want to keep it as clean as possible.
So please don't see my review as something negative, but rather something positive.

Am Mittwoch, 28. November 2012, 18:48:57 schrieb Mathias LEBLANC:

Quoted 5 lines

Ok, so i have patch the ST33 I2C driver on this branch and correct some errors. I send you the patch for the kernel 3.x I have no error on compilation, tell me if you have problems.
I have implemented the tpm_do_selftest function to get the tpm ready, but it can be removed ________________________________________

Unfortunately you attached the patch instead of sending it in plaintext which is the usual practice -> care to resend in plain text? Makes the review far easier.

(btw.: Please also have a look at
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmittingPatches;hb=HEAD
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmitChecklist;hb=HEAD
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmittingDrivers;hb=HEAD
which describes the process in detail)

When you resend the patch, can you please include the "metadata" as well - i.e. your modifications to the Kconfig / Makefile etc. I do not see a reason why to keep it in a seperate patch.

I tried the patch you've posted and it applies cleanly and now (finally) compiles as well - so now I can start with my review:

= The name =
There's already one i2c tpm driver in the tree, so maybe it would be a good idea to keep the naming scheme consistent? -> How about tpm_i2c_stm_st33.c ?
Eventually this is something Kent as a maintainer has to decice - but I would really like to see the name change. I hope we can eventually consolidate all the 'tis' based drivers.

= Compiling / License =
When compiling the driver I get the following warning
WARNING: modpost: missing MODULE_LICENSE() in /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o
Please include the appropriate MODULE_LICENSE as my kernel otherwise gets tainted by your driver.

Also this:
+ * STMicroelectronics TPM I2C Linux driver for TPM ST33ZP24
+ * Copyright (C) 2009, 2010 STMicroelectronics
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.

is not possible afaik - kernel code must be under GPL v2 _only_ without the "or (at your option) any later version." addition.

= sparse warnings =
When running sparse against your code I get the following warnings:
make -C /data/data-old/linux-2.6/ M=pwd modules C=1
make: Entering directory `/data/data-old/linux-2.6'
CHECK /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:167:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:187:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:180:5: warning: symbol 'wait_for_serirq_timeout' was not declared. Should it be static? /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:210:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:227:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:245:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:269:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:307:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:324:38: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:394:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:424:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:456:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:531:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:672:29: warning: incorrect type in assignment (different address spaces)
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:672:29: expected void [noderef] <asn:2>*iobase
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:672:29: got struct i2c_client *client
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:781:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:818:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:841:19: warning: cast removes address space of expression
CC [M] /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o
Building modules, stage 2.
MODPOST 8 modules

Please fix these if applicable - otherwise you'll probably get a friendly reminder to do so by fengguang's build test ;)

= smatch =
Same applies to smatch
make -C /data/data-old/linux-2.6/ M=pwd modules C=1 CHECK=smatch
make: Entering directory `/data/data-old/linux-2.6'
CHECK /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:535 tpm_stm_i2c_recv() warn: variable dereferenced before check 'chip' (see line 531) /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:748 tpm_st33_i2c_probe() warn: variable dereferenced before check 'platform_data' (see line 659) /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:848 tpm_st33_i2c_pm_resume() warn: should this be a bitwise op? /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:848 tpm_st33_i2c_pm_resume() warn: should this be a bitwise op?
CC [M] /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o

Please fix these if applicable - otherwise you'll probably get a friendly reminder to do so by fengguang's build test ;)

= checkpatch =
Also please run .../scripts/checkpatch.pl -strict before submission - not everything that is reported might be applicable, but quite often it is.

Quoted 7 lines

Looking forward to your v2 so I can give a more detailed code review of your code.

Thanks,
Peter

------------------------------------------------------------------------------
LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial Remotely access PCs and mobile devices and provide instant support Improve your efficiency, and focus on delivering more value-add services Discover what IT Professionals Know. Rescue delivers
http://p.sf.net/sfu/logmein_12329d2d
_______________________________________________
tpmdd-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/tpmdd-devel

#13 lore raw
Quoted 16 lines · Kent Yoder
Quoted 15 lines · Kent Yoder
Quoted 7 lines · Mathias LEBLANC

Hi Peter,

Thanks for your contribution.
I have modified the driver files name and descriptions. Regarding the warnings, it's strange.

@Kent, could you confirm that you have the same?

Yep, I see these too. They're a side effect of storing the i2c_client
pointer in vendor->iobase, which is __iomem. The i2c_client struct isn't __iomem at all though, writes in the i2c subsystem are done through client->addr. So i2c_client should really be stored somewhere else, no?

Quoted 5 lines
Quoted 3 lines

vendor->data seems like the right place. I'm on the hook to update
the other drivers with a macro for accessing vendor->data in tpm.h. I'll make these updates and send out a patch.

Mathias, one other thing, MODULE_LICENSE(): GPL?

Heh, duh, well of course it is. I've now staged everything I'm
planning on pushing at:

git://github.com/shpedoikal/linux.git tpmdd-12-05-12

Please test and let me know if I missed anything.

Thanks,
Kent

Quoted 157 lines

Kent

Quoted 130 lines

Kent

Quoted 119 lines

Regards,

Mathias Leblanc

-----Original Message-----
From: Peter Hüwe [mailto:[email protected]]
Sent: 29 November, 2012 01:05
To: Mathias LEBLANC
Cc: Kent Yoder; Kent Yoder; Jean-Luc BLANC; [email protected]; Rajiv Andrade; [email protected]; [email protected]
Subject: Re: [tpmdd-devel] [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x

Hi Mathias,

please note:
I'm writing this email on behalf of myself only and nobody else, especially not my employer - and I'm doing this in my spare time. I'm working for a direct competitor of yours, but I'm not using any knowledge that I've picked up at work or that is considered secret in any way. I have a personal interest in the TPM subsystem and want to keep it as clean as possible.
So please don't see my review as something negative, but rather something positive.

Am Mittwoch, 28. November 2012, 18:48:57 schrieb Mathias LEBLANC:

Quoted 5 lines

Ok, so i have patch the ST33 I2C driver on this branch and correct some errors. I send you the patch for the kernel 3.x I have no error on compilation, tell me if you have problems.
I have implemented the tpm_do_selftest function to get the tpm ready, but it can be removed ________________________________________

Unfortunately you attached the patch instead of sending it in plaintext which is the usual practice -> care to resend in plain text? Makes the review far easier.

(btw.: Please also have a look at
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmittingPatches;hb=HEAD
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmitChecklist;hb=HEAD
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmittingDrivers;hb=HEAD
which describes the process in detail)

When you resend the patch, can you please include the "metadata" as well - i.e. your modifications to the Kconfig / Makefile etc. I do not see a reason why to keep it in a seperate patch.

I tried the patch you've posted and it applies cleanly and now (finally) compiles as well - so now I can start with my review:

= The name =
There's already one i2c tpm driver in the tree, so maybe it would be a good idea to keep the naming scheme consistent? -> How about tpm_i2c_stm_st33.c ?
Eventually this is something Kent as a maintainer has to decice - but I would really like to see the name change. I hope we can eventually consolidate all the 'tis' based drivers.

= Compiling / License =
When compiling the driver I get the following warning
WARNING: modpost: missing MODULE_LICENSE() in /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o
Please include the appropriate MODULE_LICENSE as my kernel otherwise gets tainted by your driver.

Also this:
+ * STMicroelectronics TPM I2C Linux driver for TPM ST33ZP24
+ * Copyright (C) 2009, 2010 STMicroelectronics
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.

is not possible afaik - kernel code must be under GPL v2 _only_ without the "or (at your option) any later version." addition.

= sparse warnings =
When running sparse against your code I get the following warnings:
make -C /data/data-old/linux-2.6/ M=pwd modules C=1
make: Entering directory `/data/data-old/linux-2.6'
CHECK /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:167:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:187:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:180:5: warning: symbol 'wait_for_serirq_timeout' was not declared. Should it be static? /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:210:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:227:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:245:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:269:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:307:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:324:38: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:394:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:424:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:456:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:531:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:672:29: warning: incorrect type in assignment (different address spaces)
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:672:29: expected void [noderef] <asn:2>*iobase
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:672:29: got struct i2c_client *client
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:781:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:818:19: warning: cast removes address space of expression /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:841:19: warning: cast removes address space of expression
CC [M] /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o
Building modules, stage 2.
MODPOST 8 modules

Please fix these if applicable - otherwise you'll probably get a friendly reminder to do so by fengguang's build test ;)

= smatch =
Same applies to smatch
make -C /data/data-old/linux-2.6/ M=pwd modules C=1 CHECK=smatch
make: Entering directory `/data/data-old/linux-2.6'
CHECK /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:535 tpm_stm_i2c_recv() warn: variable dereferenced before check 'chip' (see line 531) /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:748 tpm_st33_i2c_probe() warn: variable dereferenced before check 'platform_data' (see line 659) /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:848 tpm_st33_i2c_pm_resume() warn: should this be a bitwise op? /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:848 tpm_st33_i2c_pm_resume() warn: should this be a bitwise op?
CC [M] /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o

Please fix these if applicable - otherwise you'll probably get a friendly reminder to do so by fengguang's build test ;)

= checkpatch =
Also please run .../scripts/checkpatch.pl -strict before submission - not everything that is reported might be applicable, but quite often it is.

Quoted 7 lines

Looking forward to your v2 so I can give a more detailed code review of your code.

Thanks,
Peter

Quoted 11 lines

------------------------------------------------------------------------------
LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial Remotely access PCs and mobile devices and provide instant support Improve your efficiency, and focus on delivering more value-add services Discover what IT Professionals Know. Rescue delivers
http://p.sf.net/sfu/logmein_12329d2d
_______________________________________________
tpmdd-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/tpmdd-devel

------------------------------------------------------------------------------
LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial Remotely access PCs and mobile devices and provide instant support Improve your efficiency, and focus on delivering more value-add services Discover what IT Professionals Know. Rescue delivers
http://p.sf.net/sfu/logmein_12329d2d
_______________________________________________
tpmdd-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/tpmdd-devel

maintainer
#14 lore raw

Hi Kent,

Quoted 4 lines

Heh, duh, well of course it is. I've now staged everything I'm
planning on pushing at:

Quoted 9 lines

git://github.com/shpedoikal/linux.git tpmdd-12-05-12

Please test and let me know if I missed anything.

Thanks,
Kent

Quoted 2 lines

Kent

(I'm still writing this on behalf of myself ;)

<pedantic mode on - please don't take it personally>
While I really appreciate you helping Mathias out here, I'm not so sure whether an offlist discussion of a driver submission is a really good idea. I did not see any v2 / improvements on list and now it's commited (?!)

There's no need to argue here, I'm fine with this,
but I'd highly appreciate if at least the result is published again in the future on the mailing list, for proper review.

If the mailfilter got the mails I apologize.
<pedantic mode off>

Can you perhaps post the message to the list, so a proper review is possible?

I thought the Mathias has changed the naming of the files? I don't see that in your commit.

The version you committed to that branch still has some of the items of my first review. (e.g.
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.)

-> The (possible) GPL v3 clause has to go away for the kernel, but I'm not a lawyer.

Thanks,
Peter

#15 lore raw
Quoted 15 lines · Peter Hüwe

Hi Kent,

Quoted 4 lines

Heh, duh, well of course it is. I've now staged everything I'm
planning on pushing at:

Quoted 9 lines

git://github.com/shpedoikal/linux.git tpmdd-12-05-12

Please test and let me know if I missed anything.

Thanks,
Kent

Quoted 2 lines

Kent

Quoted 6 lines

(I'm still writing this on behalf of myself ;)

<pedantic mode on - please don't take it personally>
While I really appreciate you helping Mathias out here, I'm not so sure whether an offlist discussion of a driver submission is a really good idea. I did not see any v2 / improvements on list and now it's commited (?!)

This is already at least v3 of the driver IIRC. And its not fully
committed at all, its just in a staging tree. I can blow it away at any time if we find problems.

Quoted 3 lines

There's no need to argue here, I'm fine with this,
but I'd highly appreciate if at least the result is published again in the future on the mailing list, for proper review.

Not a problem. usually I'd attach any updates I planned to commit for
public review, then if anything wasn't made public before I issue my pull request to send them to security-next, I'd append the full diff to the pull request at that time. Either way, all changes will hit a list at some point in time.

Quoted 7 lines

If the mailfilter got the mails I apologize.
<pedantic mode off>

Can you perhaps post the message to the list, so a proper review is possible?

I thought the Mathias has changed the naming of the files? I don't see that in your commit.

Good catch, this is the kind of review I was asking for. :-)

Quoted 10 lines

The version you committed to that branch still has some of the items of my first review. (e.g.
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.)

-> The (possible) GPL v3 clause has to go away for the kernel, but I'm not a lawyer.

A GPLv3 clause would say "v3 or any later version". This should be
fine.

Kent

Quoted 5 lines

Thanks,
Peter

maintainer
#16 lore raw

Hi Kent,

Am Mittwoch, 5. Dezember 2012, 22:00:48 schrieb Kent Yoder:

Quoted 9 lines

This is already at least v3 of the driver IIRC. And its not fully
committed at all, its just in a staging tree. I can blow it away at any time if we find problems.

Not a problem. usually I'd attach any updates I planned to commit for
public review, then if anything wasn't made public before I issue my pull request to send them to security-next, I'd append the full diff to the pull request at that time. Either way, all changes will hit a list at some point in time.

Ok - great ;) maybe my I got the 'planning on pushing' wrong.

Quoted 4 lines
Quoted 2 lines

I thought the Mathias has changed the naming of the files? I don't see that in your commit.

Good catch, this is the kind of review I was asking for. :-)

Hehe. Glad to hear.
You know I care for the tpm subsystem ;)

Quoted 10 lines
Quoted 10 lines

The version you committed to that branch still has some of the items of my first review. (e.g.

* This program is free software; you can redistribute it and/or modify

* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.)

Quoted 5 lines
Quoted 2 lines

-> The (possible) GPL v3 clause has to go away for the kernel, but I'm not a lawyer.

A GPLv3 clause would say "v3 or any later version". This should be
fine.

Hmm, okay.
I just googled for it on lkml, there are some files and drivers. So I guess it's okay - sorry for the noise.
But I'm not a lawyer, personally I'd go for GPLv2 only but that's my personal opinion.

Thanks,
Peter

maintainer
#17 lore raw
In reply to Kent Yoder

Hi Kent, Matthias,

Am Mittwoch, 5. Dezember 2012, 19:07:20 schrieb Kent Yoder:

Quoted 6 lines

Heh, duh, well of course it is. I've now staged everything I'm
planning on pushing at:

git://github.com/shpedoikal/linux.git tpmdd-12-05-12

Please test and let me know if I missed anything.

I just checked out your commit from github.

Here's a part of the review:

Smatch still complains a bit, sparse is fine.:

make -C /data/data-old/linux-2.6/ M=pwd C=1 CHECK=smatch modules
make: Entering directory `/data/data-old/linux-2.6'
CHECK /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:535 tpm_stm_i2c_recv() warn: variable dereferenced before check 'chip' (see line 531) /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:748 tpm_st33_i2c_probe() warn: variable dereferenced before check 'platform_data' (see line 659) /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:848 tpm_st33_i2c_pm_resume() warn: should this be a bitwise op? /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:848 tpm_st33_i2c_pm_resume() warn: should this be a bitwise op?
CC [M] /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o
Building modules, stage 2.
MODPOST 6 modules
LD [M] /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.ko
make: Leaving directory `/data/data-old/linux-2.6'

Please check.

(maybe also fix the checkpatch -strict stuff as well? Would be nice ;)

drivers/char/tpm/tpm_stm_st33_i2c.h
- Does the driver really need a seperate headerfile in drivers/char/tpm/ ? for me it seems everything can be included into the c-file.

Quoted 4 lines
struct st_tpm_hash {
	int size;
	u8 *data;
};

is unused - please remove.

Quoted 1 line

#define MINOR_NUM_I2C 224

Please remove, it's unused and if you really need it use the one from tpm.h

include/linux/i2c/tpm_stm_st33_i2c.h
I'm not sure if this is needed publicly? Or does only your driver need this?

Quoted 1 line

struct st33zp24_platform_data {

Telling from the name I have no idea what this device is.

drivers/char/tpm/tpm_stm_st33_i2c.c

Quoted 9 lines
enum stm33zp24_int_flags {
	TPM_GLOBAL_INT_ENABLE = 0x80,
	TPM_INTF_CMD_READY_INT = 0x080,
	TPM_INTF_FIFO_AVALAIBLE_INT = 0x040,
	TPM_INTF_WAKE_UP_READY_INT = 0x020,
	TPM_INTF_LOCALITY_CHANGE_INT = 0x004,
	TPM_INTF_STS_VALID_INT = 0x002,
	TPM_INTF_DATA_AVAIL_INT = 0x001,
};

Why the leading zeros? please remove.

Quoted 4 lines

static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
...
static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
,,,

Quoted 11 lines
static struct i2c_driver tpm_st33_i2c_driver = {
	.driver = {
		   .owner = THIS_MODULE,
		   .name = TPM_ST33_I2C,
		   },
	.probe = tpm_st33_i2c_probe,
	.remove = tpm_st33_i2c_remove,
	.resume = tpm_st33_i2c_pm_resume,
	.suspend = tpm_st33_i2c_pm_suspend,
	.id_table = tpm_st33_i2c_id
};

Please convert resume/suspend to .driver.pm

It's pretty easy.
See this post for details
http://sourceforge.net/mailarchive/message.php?msg_id=29516784
Rafael did spent quite a lot of effort to convert almost every driver back then, so we should 'fix' new ones.

Quoted 14 lines

/*
* tpm_st33_i2c_init initialize driver
* @return: 0 if successful, else non zero value.
*/
static int __init tpm_st33_i2c_init(void)

{
	return i2c_add_driver(&tpm_st33_i2c_driver);
}

/*
* tpm_st33_i2c_exit The kernel calls this function during unloading the
* module or during shut down process
*/
static void __exit tpm_st33_i2c_exit(void)

Quoted 6 lines
{
	i2c_del_driver(&tpm_st33_i2c_driver);
}

module_init(tpm_st33_i2c_init);
module_exit(tpm_st33_i2c_exit);

Hooray for oneliners ;)
+ module_i2c_driver(tpm_st33_i2c_driver);

Keep on hacking ;)

Thanks,
PeterH

#18 lore raw
Quoted 92 lines · Peter Hüwe

Hi Kent, Matthias,

Am Mittwoch, 5. Dezember 2012, 19:07:20 schrieb Kent Yoder:

Quoted 6 lines

Heh, duh, well of course it is. I've now staged everything I'm
planning on pushing at:

git://github.com/shpedoikal/linux.git tpmdd-12-05-12

Please test and let me know if I missed anything.

I just checked out your commit from github.

Here's a part of the review:

Smatch still complains a bit, sparse is fine.:

make -C /data/data-old/linux-2.6/ M=pwd C=1 CHECK=smatch modules
make: Entering directory `/data/data-old/linux-2.6'
CHECK /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c
/data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:535 tpm_stm_i2c_recv() warn: variable dereferenced before check 'chip' (see line 531) /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:748 tpm_st33_i2c_probe() warn: variable dereferenced before check 'platform_data' (see line 659) /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:848 tpm_st33_i2c_pm_resume() warn: should this be a bitwise op? /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.c:848 tpm_st33_i2c_pm_resume() warn: should this be a bitwise op?
CC [M] /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.o
Building modules, stage 2.
MODPOST 6 modules
LD [M] /data/data-old/linux-2.6/drivers/char/tpm/tpm_stm_st33_i2c.ko
make: Leaving directory `/data/data-old/linux-2.6'

Please check.

(maybe also fix the checkpatch -strict stuff as well? Would be nice ;)

drivers/char/tpm/tpm_stm_st33_i2c.h
- Does the driver really need a seperate headerfile in drivers/char/tpm/ ? for me it seems everything can be included into the c-file.

Quoted 4 lines
struct st_tpm_hash {
	int size;
	u8 *data;
};

is unused - please remove.

Quoted 1 line

#define MINOR_NUM_I2C 224

Please remove, it's unused and if you really need it use the one from tpm.h

include/linux/i2c/tpm_stm_st33_i2c.h
I'm not sure if this is needed publicly? Or does only your driver need this?

Quoted 1 line

struct st33zp24_platform_data {

Telling from the name I have no idea what this device is.

drivers/char/tpm/tpm_stm_st33_i2c.c

Quoted 9 lines
enum stm33zp24_int_flags {
	TPM_GLOBAL_INT_ENABLE = 0x80,
	TPM_INTF_CMD_READY_INT = 0x080,
	TPM_INTF_FIFO_AVALAIBLE_INT = 0x040,
	TPM_INTF_WAKE_UP_READY_INT = 0x020,
	TPM_INTF_LOCALITY_CHANGE_INT = 0x004,
	TPM_INTF_STS_VALID_INT = 0x002,
	TPM_INTF_DATA_AVAIL_INT = 0x001,
};

Why the leading zeros? please remove.

Quoted 4 lines

static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
...
static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
,,,

Quoted 11 lines
static struct i2c_driver tpm_st33_i2c_driver = {
	.driver = {
		   .owner = THIS_MODULE,
		   .name = TPM_ST33_I2C,
		   },
	.probe = tpm_st33_i2c_probe,
	.remove = tpm_st33_i2c_remove,
	.resume = tpm_st33_i2c_pm_resume,
	.suspend = tpm_st33_i2c_pm_suspend,
	.id_table = tpm_st33_i2c_id
};
Quoted 6 lines

Please convert resume/suspend to .driver.pm

It's pretty easy.
See this post for details
http://sourceforge.net/mailarchive/message.php?msg_id=29516784
Rafael did spent quite a lot of effort to convert almost every driver back then, so we should 'fix' new ones.

Not sure how easy this will be considering these routines are i2c-specific -- they don't just call the tpm_tpm_* functions like the other drivers.

Anyway, I left this off but I believe I've fixed up the other major issues. We can leave the more nit-picky stuff for another day.

Please test:

git://github.com/shpedoikal/linux.git tpmdd-12-06-12

Kent

Quoted 34 lines
Quoted 16 lines

/*
* tpm_st33_i2c_init initialize driver
* @return: 0 if successful, else non zero value.
*/
static int __init tpm_st33_i2c_init(void)

{
	return i2c_add_driver(&tpm_st33_i2c_driver);
}

/*
* tpm_st33_i2c_exit The kernel calls this function during unloading the
* module or during shut down process
*/
static void __exit tpm_st33_i2c_exit(void)

Quoted 6 lines
{
	i2c_del_driver(&tpm_st33_i2c_driver);
}

module_init(tpm_st33_i2c_init);
module_exit(tpm_st33_i2c_exit);

Hooray for oneliners ;)
+ module_i2c_driver(tpm_st33_i2c_driver);

Keep on hacking ;)

Thanks,
PeterH

maintainer
#19 lore raw

Am Donnerstag, 6. Dezember 2012, 00:09:41 schrieb Kent Yoder:

Quoted 33 lines
Quoted 25 lines
Quoted 23 lines

static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)

...

static int tpm_st33_i2c_pm_resume(struct i2c_client *client)

,,,
static struct i2c_driver tpm_st33_i2c_driver = {

	.driver = {
	
		   .owner = THIS_MODULE,
		   .name = TPM_ST33_I2C,
		   },
	
	.probe = tpm_st33_i2c_probe,
	.remove = tpm_st33_i2c_remove,
	.resume = tpm_st33_i2c_pm_resume,
	.suspend = tpm_st33_i2c_pm_suspend,
	.id_table = tpm_st33_i2c_id
Quoted 1 line

};

Quoted 7 lines

Please convert resume/suspend to .driver.pm

It's pretty easy.
See this post for details
http://sourceforge.net/mailarchive/message.php?msg_id=29516784
Rafael did spent quite a lot of effort to convert almost every driver back then, so we should 'fix' new ones.

Quoted 3 lines

Not sure how easy this will be considering these routines are i2c-specific -- they don't just call the tpm_tpm_* functions like the other drivers.

Why are they really i2c-specific?
The only i2c specific part is the i2c_get_clientdata - which calls get_dev_drvdata internally.
And I since we already get the correct struct device we don't need to use TPM_VPRIV in order to access the platform data.

I'll send a patch - see next email.
@Matthias: can you please test this? As I cannot do this.

Thanks,
Peter

#20 lore raw
Quoted 38 lines · Peter Hüwe

Am Donnerstag, 6. Dezember 2012, 00:09:41 schrieb Kent Yoder:

Quoted 33 lines
Quoted 25 lines
Quoted 23 lines

static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)

...

static int tpm_st33_i2c_pm_resume(struct i2c_client *client)

,,,
static struct i2c_driver tpm_st33_i2c_driver = {

	.driver = {
	
		   .owner = THIS_MODULE,
		   .name = TPM_ST33_I2C,
		   },
	
	.probe = tpm_st33_i2c_probe,
	.remove = tpm_st33_i2c_remove,
	.resume = tpm_st33_i2c_pm_resume,
	.suspend = tpm_st33_i2c_pm_suspend,
	.id_table = tpm_st33_i2c_id
Quoted 1 line

};

Quoted 7 lines

Please convert resume/suspend to .driver.pm

It's pretty easy.
See this post for details
http://sourceforge.net/mailarchive/message.php?msg_id=29516784
Rafael did spent quite a lot of effort to convert almost every driver back then, so we should 'fix' new ones.

Quoted 3 lines

Not sure how easy this will be considering these routines are i2c-specific -- they don't just call the tpm_tpm_* functions like the other drivers.

Quoted 5 lines

Why are they really i2c-specific?
The only i2c specific part is the i2c_get_clientdata - which calls get_dev_drvdata internally.
And I since we already get the correct struct device we don't need to use TPM_VPRIV in order to access the platform data.

I was referring to the power mgmt stuff. Those functions take struct
device, not a i2c_client *.

Kent

Quoted 8 lines

I'll send a patch - see next email.
@Matthias: can you please test this? As I cannot do this.

Thanks,
Peter

maintainer
#21 lore raw

Re: [tpmdd-devel] [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x OOPS!

Hi,

since I don't have a tpm I simply tried out whether the handling without a tpm is correct, which it unfortunately isn't as my kernel oopses.

Steps to reproduce:

# insmod  /data/data-old/linux-2.6/drivers/char/tpm/tpm_i2c_stm_st33.ko
# echo st33zp24_i2c 0x20 /sys/bus/i2c/devices/i2c-1/new_device 
BUG: unable to handle kernel NULL pointer dereference at 0000000000000018
IP: [<ffffffffa02ff4ad>] tpm_st33_i2c_probe+0x87/0x21a [tpm_i2c_stm_st33]
PGD 3dcbbe067 PUD 3e994b067 PMD 0
Oops: 0002 [#1] PREEMPT SMP
Modules linked in: tpm_i2c_stm_st33(O) tpm(O) tpm_bios(O) w83627ehf hwmon_vid 

ipv6 snd_hda_codec_hdmi snd_hda_codec_realtek joydev usbhid coretemp kvm_intel kvm ghash_clmulni_intel sg pcspkr snd_hda_intel snd_hda_codec snd_hwdep snd_pcm snd_timer snd i2c_i801 snd_page_alloc mei [last unloaded:
tpm_i2c_stm_st33]
CPU 2
Pid: 4454, comm: bash Tainted: G W O 3.6.9 #14 To Be Filled By O.E.M.
To Be Filled By O.E.M./Z77 Pro4

RIP: 0010:[<ffffffffa02ff4ad>]  [<ffffffffa02ff4ad>] tpm_st33_i2c_probe+0x87/0x21a 
[tpm_i2c_stm_st33]
RSP: 0018:ffff8803dcb87c18  EFLAGS: 00010286
RAX: ffff8803eb0d1800 RBX: ffff8803ff710400 RCX: 0000000000000000
RDX: 0000000000000000 RSI: ffff88041d0026d0 RDI: 0000000000000001
RBP: ffff8803dcb87c38 R08: 000000000000000b R09: 0000000000000001
R10: 0000000000000001 R11: ffff8803ff710400 R12: 0000000000000000
R13: ffff8803ff710c00 R14: ffff8803ff710c28 R15: ffff8803ff710c04
FS:  00007f520567a700(0000) GS:ffff88041f280000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000018 CR3: 00000003e48b0000 CR4: 00000000001407e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process bash (pid: 4454, threadinfo ffff8803dcb86000, task ffff8803c87226c0)
Stack:
 ffff8803ff710c28 ffffffffa02ffb80 ffffffffa02ff426 ffff8803ff710c00
 ffff8803dcb87c78 ffffffff8134079a ffffffffa02ffe00 ffff8803ff710c28
 00000000ffffffed ffffffffa02ffe00 0000000000000000 ffff88040ebe4460
Call Trace:
 [<ffffffffa02ff426>] ? request_locality+0xc2/0xc2 [tpm_i2c_stm_st33]
 [<ffffffff8134079a>] i2c_device_probe+0xa4/0xcd
 [<ffffffff812b9c24>] driver_probe_device+0xa9/0x1bf
 [<ffffffff812b9de9>] __device_attach+0x35/0x3a
 [<ffffffff812b9db4>] ? __driver_attach+0x7a/0x7a
 [<ffffffff812b84d7>] bus_for_each_drv+0x51/0x87
 [<ffffffff812b9b3f>] device_attach+0x6e/0x8e
 [<ffffffff812b91b0>] bus_probe_device+0x2d/0x98
 [<ffffffff812b7ace>] device_add+0x3e1/0x548
 [<ffffffff812bf2b3>] ? device_pm_init+0x60/0x84
 [<ffffffff812b7c4c>] device_register+0x17/0x1b
 [<ffffffff813410c4>] i2c_new_device+0x12b/0x175
 [<ffffffff813411e4>] i2c_sysfs_new_device+0xd6/0x15a
 [<ffffffff812b6394>] dev_attr_store+0x1b/0x1d
 [<ffffffff81136f39>] sysfs_write_file+0xef/0x12b
 [<ffffffff810e4ac3>] vfs_write+0xa9/0x119
 [<ffffffff810e4cfa>] sys_write+0x45/0x6c
 [<ffffffff814005a2>] system_call_fastpath+0x16/0x1b
Code: ff 48 c7 c6 a7 fa 2f a0 48 85 c0 48 89 c3 74 cd 48 8b 3d d7 d0 2a e1 be 

d0 00 00 00 4d 8b a5 c8 00 00 00 e8 3b 32 de e0 48 85 c0 <49> 89 44 24 18 0f 84 5a 01 00 00 48 8b 3d b1 d0 2a e1 be d0 00
RIP [<ffffffffa02ff4ad>] tpm_st33_i2c_probe+0x87/0x21a [tpm_i2c_stm_st33]
RSP <ffff8803dcb87c18>
CR2: 0000000000000018

---[ end trace ddc5676681e8ed72 ]---

Thanks,
Peter

A message that is not in the archive (sent off-list or not yet received): [email protected]

#23 lore raw

RE: [tpmdd-devel] [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x OOPS!

Hi Peter,

Do you have the I2C configuration activated ?
I have this error when I add the driver in a kernel with no I2C activated

-----Original Message-----
From: Kent Yoder [mailto:[email protected]]
Sent: 08 January, 2013 18:21
To: Mathias LEBLANC
Cc: Peter Hüwe; Jean-Luc BLANC; [email protected]; [email protected]; [email protected]; Kent Yoder
Subject: Re: [tpmdd-devel] [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x OOPS!

Quoted 9 lines · Peter Hüwe

Hi,

since I don't have a tpm I simply tried out whether the handling without a tpm is correct, which it unfortunately isn't as my kernel oopses.

Steps to reproduce:

# insmod  
/data/data-old/linux-2.6/drivers/char/tpm/tpm_i2c_stm_st33.ko
# echo st33zp24_i2c 0x20 /sys/bus/i2c/devices/i2c-1/new_device

Mathias, can you address this oops? We can't merge this driver until its resolved.

Kent

Quoted 69 lines
BUG: unable to handle kernel NULL pointer dereference at 
0000000000000018
IP: [<ffffffffa02ff4ad>] tpm_st33_i2c_probe+0x87/0x21a 
[tpm_i2c_stm_st33] PGD 3dcbbe067 PUD 3e994b067 PMD 0
Oops: 0002 [#1] PREEMPT SMP
Modules linked in: tpm_i2c_stm_st33(O) tpm(O) tpm_bios(O) w83627ehf 

hwmon_vid
ipv6 snd_hda_codec_hdmi snd_hda_codec_realtek joydev usbhid coretemp kvm_intel kvm ghash_clmulni_intel sg pcspkr snd_hda_intel
snd_hda_codec snd_hwdep snd_pcm snd_timer snd i2c_i801 snd_page_alloc mei [last unloaded:
tpm_i2c_stm_st33]
CPU 2
Pid: 4454, comm: bash Tainted: G W O 3.6.9 #14 To Be Filled By O.E.M.
To Be Filled By O.E.M./Z77 Pro4

RIP: 0010:[<ffffffffa02ff4ad>]  [<ffffffffa02ff4ad>] 
tpm_st33_i2c_probe+0x87/0x21a [tpm_i2c_stm_st33]
RSP: 0018:ffff8803dcb87c18  EFLAGS: 00010286
RAX: ffff8803eb0d1800 RBX: ffff8803ff710400 RCX: 0000000000000000
RDX: 0000000000000000 RSI: ffff88041d0026d0 RDI: 0000000000000001
RBP: ffff8803dcb87c38 R08: 000000000000000b R09: 0000000000000001
R10: 0000000000000001 R11: ffff8803ff710400 R12: 0000000000000000
R13: ffff8803ff710c00 R14: ffff8803ff710c28 R15: ffff8803ff710c04
FS:  00007f520567a700(0000) GS:ffff88041f280000(0000) 

knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000018 CR3: 00000003e48b0000 CR4: 00000000001407e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process bash (pid: 4454, threadinfo ffff8803dcb86000, task ffff8803c87226c0)

Stack:
 ffff8803ff710c28 ffffffffa02ffb80 ffffffffa02ff426 ffff8803ff710c00
 ffff8803dcb87c78 ffffffff8134079a ffffffffa02ffe00 ffff8803ff710c28  
00000000ffffffed ffffffffa02ffe00 0000000000000000 ffff88040ebe4460 
Call Trace:
 [<ffffffffa02ff426>] ? request_locality+0xc2/0xc2 [tpm_i2c_stm_st33]  
[<ffffffff8134079a>] i2c_device_probe+0xa4/0xcd  [<ffffffff812b9c24>] 
driver_probe_device+0xa9/0x1bf  [<ffffffff812b9de9>] 
__device_attach+0x35/0x3a  [<ffffffff812b9db4>] ? 
__driver_attach+0x7a/0x7a  [<ffffffff812b84d7>] 
bus_for_each_drv+0x51/0x87  [<ffffffff812b9b3f>] 
device_attach+0x6e/0x8e  [<ffffffff812b91b0>] 
bus_probe_device+0x2d/0x98  [<ffffffff812b7ace>] 
device_add+0x3e1/0x548  [<ffffffff812bf2b3>] ? 
device_pm_init+0x60/0x84  [<ffffffff812b7c4c>] 
device_register+0x17/0x1b  [<ffffffff813410c4>] 
i2c_new_device+0x12b/0x175  [<ffffffff813411e4>] 
i2c_sysfs_new_device+0xd6/0x15a  [<ffffffff812b6394>] 
dev_attr_store+0x1b/0x1d  [<ffffffff81136f39>] 
sysfs_write_file+0xef/0x12b  [<ffffffff810e4ac3>] vfs_write+0xa9/0x119  
[<ffffffff810e4cfa>] sys_write+0x45/0x6c  [<ffffffff814005a2>] 
system_call_fastpath+0x16/0x1b
Code: ff 48 c7 c6 a7 fa 2f a0 48 85 c0 48 89 c3 74 cd 48 8b 3d d7 d0 

2a e1 be
d0 00 00 00 4d 8b a5 c8 00 00 00 e8 3b 32 de e0 48 85 c0 <49> 89 44 24 18 0f
84 5a 01 00 00 48 8b 3d b1 d0 2a e1 be d0 00 RIP [<ffffffffa02ff4ad>] tpm_st33_i2c_probe+0x87/0x21a [tpm_i2c_stm_st33] RSP <ffff8803dcb87c18>
CR2: 0000000000000018

---[ end trace ddc5676681e8ed72 ]---

Thanks,
Peter

maintainer
#24 lore raw

Re: [tpmdd-devel] [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x OOPS!

Hi,

Quoted 2 lines

Do you have the I2C configuration activated ?
I have this error when I add the driver in a kernel with no I2C activated

I2C works fine on my beaglebone, however I did not add the platform data to my boardfile.
a) as I don't use board files anymore but device trees ;)
b) it's perfectly legit to load a driver without the platform data and probe it at runtime (which then in doubt fails to probe).

In my opinion a driver should never ever oops, even if the user does something stupid (e.g. like probing from userspace ;)

The fix is as simple as adding a simple
if (!platform_data) return -ENODEV;
or something similar.

Thanks,
Peter

#25 lore raw

Re: [tpmdd-devel] [PATCH 1/1] TPM: STMicroelectronics ST33 I2C KERNEL 3.x.x OOPS!

Quoted 2 lines · Peter Hüwe

In my opinion a driver should never ever oops, even if the user does something stupid (e.g. like probing from userspace ;)

Exactly... I've coded up a fix for the oops and staged it here, please test:

https://github.com/shpedoikal/linux.git tpmdd-01-22-13

Thanks,
Kent

Quoted 7 lines

The fix is as simple as adding a simple
if (!platform_data) return -ENODEV;
or something similar.

Thanks,
Peter

maintainer
#26 lore raw

PATCH 1/1 char/tpm: Use struct dev_pm_ops for power management

In reply to Kent Yoder

This patch converts the suspend and resume functions for tpm_i2c_stm_st33 to the new dev_pm_ops.

---
Note:
I'm sending this patch on behalf of myself only and nobody else,
especially not my employer - and I'm doing this in my spare time.
I have to inform you that I'm working for a direct competitor of the driver author, but I'm not using any knowledge that is considered secret in any way.
I have a personal interest in the TPM subsystem and want to keep it as clean as possible.

 drivers/char/tpm/tpm_i2c_stm_st33.c |   25 ++++++++++---------------
 1 files changed, 10 insertions(+), 15 deletions(-)
drivers/char/tpm/tpm_i2c_stm_st33.c +10 -15
index b11201f..ecbb81e 100644
@@ -826,12 +826,10 @@ static __devexit int tpm_st33_i2c_remove(struct i2c_client *client)
  * @param: mesg, the power management message.
  * @return: 0 in case of success.
  */
-static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
+static int tpm_st33_i2c_pm_suspend(struct device *dev)
 {
-	struct tpm_chip *chip =
-	    (struct tpm_chip *)i2c_get_clientdata(client);
-	struct st33zp24_platform_data *pin_infos =
-		((struct i2c_client *)TPM_VPRIV(chip))->dev.platform_data;
+	struct tpm_chip *chip = dev_get_drvdata(dev);
+	struct st33zp24_platform_data *pin_infos = dev->platform_data;
 	int ret = 0;
 
 	if (power_mgt)
@@ -839,7 +837,7 @@ static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
 	else{
 		if (chip->data_buffer == NULL)
 			chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-		ret = tpm_pm_suspend(&client->dev);
+		ret = tpm_pm_suspend(dev);
 	}
 	return ret;
 }				/* tpm_st33_i2c_suspend() */
@@ -849,12 +847,10 @@ static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
  * @param: client, the i2c_client drescription (TPM I2C description).
  * @return: 0 in case of success.
  */
-static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
+static int tpm_st33_i2c_pm_resume(struct device *dev)
 {
-	struct tpm_chip *chip =
-	    (struct tpm_chip *)i2c_get_clientdata(client);
-	struct st33zp24_platform_data *pin_infos =
-		((struct i2c_client *)TPM_VPRIV(chip))->dev.platform_data;
+	struct tpm_chip *chip = dev_get_drvdata(dev);
+	struct st33zp24_platform_data *pin_infos = dev->platform_data;
 
 	int ret = 0;
 
@@ -867,7 +863,7 @@ static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
 	} else{
 	if (chip->data_buffer == NULL)
 		chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-	ret = tpm_pm_resume(&client->dev);
+	ret = tpm_pm_resume(dev);
 	if (!ret)
 		tpm_do_selftest(chip);
 	}
@@ -880,16 +876,15 @@ static const struct i2c_device_id tpm_st33_i2c_id[] = {
 };
 
 MODULE_DEVICE_TABLE(i2c, tpm_st33_i2c_id);
-
+static SIMPLE_DEV_PM_OPS(tpm_st33_i2c_ops, tpm_st33_i2c_pm_suspend, tpm_st33_i2c_pm_resume);
 static struct i2c_driver tpm_st33_i2c_driver = {
 	.driver = {
 		   .owner = THIS_MODULE,
 		   .name = TPM_ST33_I2C,
+		   .pm = &tpm_st33_i2c_ops,
 		   },
 	.probe = tpm_st33_i2c_probe,
 	.remove = tpm_st33_i2c_remove,
-	.resume = tpm_st33_i2c_pm_resume,
-	.suspend = tpm_st33_i2c_pm_suspend,
 	.id_table = tpm_st33_i2c_id
 };
 
Signature · 1 line
1.7.8.6
#27 lore raw

Re: [PATCH] char/tpm: Use struct dev_pm_ops for power management

Quoted 2 lines · Peter Huewe

This patch converts the suspend and resume functions for tpm_i2c_stm_st33 to the new dev_pm_ops.

Ignore my last mail. I'll take a look at this.

Thanks,
Kent

Quoted 89 lines

---
Note:
I'm sending this patch on behalf of myself only and nobody else,
especially not my employer - and I'm doing this in my spare time.
I have to inform you that I'm working for a direct competitor of the driver author, but I'm not using any knowledge that is considered secret in any way.
I have a personal interest in the TPM subsystem and want to keep it as clean as possible.

 drivers/char/tpm/tpm_i2c_stm_st33.c |   25 ++++++++++---------------
 1 files changed, 10 insertions(+), 15 deletions(-)
drivers/char/tpm/tpm_i2c_stm_st33.c +10 -15
index b11201f..ecbb81e 100644
@@ -826,12 +826,10 @@ static __devexit int tpm_st33_i2c_remove(struct i2c_client *client)
  * @param: mesg, the power management message.
  * @return: 0 in case of success.
  */
-static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
+static int tpm_st33_i2c_pm_suspend(struct device *dev)
 {
-	struct tpm_chip *chip =
-	    (struct tpm_chip *)i2c_get_clientdata(client);
-	struct st33zp24_platform_data *pin_infos =
-		((struct i2c_client *)TPM_VPRIV(chip))->dev.platform_data;
+	struct tpm_chip *chip = dev_get_drvdata(dev);
+	struct st33zp24_platform_data *pin_infos = dev->platform_data;
 	int ret = 0;
 
 	if (power_mgt)
@@ -839,7 +837,7 @@ static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
 	else{
 		if (chip->data_buffer == NULL)
 			chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-		ret = tpm_pm_suspend(&client->dev);
+		ret = tpm_pm_suspend(dev);
 	}
 	return ret;
 }				/* tpm_st33_i2c_suspend() */
@@ -849,12 +847,10 @@ static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
  * @param: client, the i2c_client drescription (TPM I2C description).
  * @return: 0 in case of success.
  */
-static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
+static int tpm_st33_i2c_pm_resume(struct device *dev)
 {
-	struct tpm_chip *chip =
-	    (struct tpm_chip *)i2c_get_clientdata(client);
-	struct st33zp24_platform_data *pin_infos =
-		((struct i2c_client *)TPM_VPRIV(chip))->dev.platform_data;
+	struct tpm_chip *chip = dev_get_drvdata(dev);
+	struct st33zp24_platform_data *pin_infos = dev->platform_data;
 
 	int ret = 0;
 
@@ -867,7 +863,7 @@ static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
 	} else{
 	if (chip->data_buffer == NULL)
 		chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-	ret = tpm_pm_resume(&client->dev);
+	ret = tpm_pm_resume(dev);
 	if (!ret)
 		tpm_do_selftest(chip);
 	}
@@ -880,16 +876,15 @@ static const struct i2c_device_id tpm_st33_i2c_id[] = {
 };
 
 MODULE_DEVICE_TABLE(i2c, tpm_st33_i2c_id);
-
+static SIMPLE_DEV_PM_OPS(tpm_st33_i2c_ops, tpm_st33_i2c_pm_suspend, tpm_st33_i2c_pm_resume);
 static struct i2c_driver tpm_st33_i2c_driver = {
 	.driver = {
 		   .owner = THIS_MODULE,
 		   .name = TPM_ST33_I2C,
+		   .pm = &tpm_st33_i2c_ops,
 		   },
 	.probe = tpm_st33_i2c_probe,
 	.remove = tpm_st33_i2c_remove,
-	.resume = tpm_st33_i2c_pm_resume,
-	.suspend = tpm_st33_i2c_pm_suspend,
 	.id_table = tpm_st33_i2c_id
 };
 
Signature · 1 line
1.7.8.6
#28 lore raw

RE: [PATCH] char/tpm: Use struct dev_pm_ops for power management

Hi Kent, Peter,

I am never in my office Thursday and Friday, I just saw your emails today. I will transfer the tpm_i2c_stm_st33.h code to tpm_i2c_stm_st33.c
@Peter: I tested your patch and activate the CONFIG_PM_SLEEP in the kernel compilation. Thanks Peter!

About:
"Not sure how easy this will be considering these routines are

Quoted 2 lines
Quoted 2 lines

i2c-specific -- they don't just call the tpm_tpm_* functions like the other drivers."

We have created another function similar to the tpm_* functions, to ensure the good work of the tpm on each platform for each case. But anyway , we use the tpm_* function in them.

Thanks,
Mathias Leblanc

-----Original Message-----
From: Kent Yoder [mailto:[email protected]]
Sent: 06 December, 2012 16:08
To: Peter Huewe
Cc: Mathias LEBLANC; Jean-Luc BLANC; [email protected]; [email protected]; Rajiv Andrade; [email protected]; Kent Yoder
Subject: Re: [PATCH] char/tpm: Use struct dev_pm_ops for power management

Quoted 2 lines · Peter Huewe

This patch converts the suspend and resume functions for tpm_i2c_stm_st33 to the new dev_pm_ops.

Ignore my last mail. I'll take a look at this.

Thanks,
Kent

Quoted 92 lines

---
Note:
I'm sending this patch on behalf of myself only and nobody else,
especially not my employer - and I'm doing this in my spare time.
I have to inform you that I'm working for a direct competitor of the
driver author, but I'm not using any knowledge that is considered secret in any way. I have a personal interest in the TPM subsystem and want to keep it as clean as possible.

 drivers/char/tpm/tpm_i2c_stm_st33.c |   25 ++++++++++---------------
 1 files changed, 10 insertions(+), 15 deletions(-)
drivers/char/tpm/tpm_i2c_stm_st33.c +3 -5
index b11201f..ecbb81e 100644
@@ -826,12 +826,10 @@ static __devexit int tpm_st33_i2c_remove(struct i2c_client *client)
  * @param: mesg, the power management message.
  * @return: 0 in case of success.
  */
-static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, 
 pm_message_t mesg)
+static int tpm_st33_i2c_pm_suspend(struct device *dev)
 {
-	struct tpm_chip *chip =
-	    (struct tpm_chip *)i2c_get_clientdata(client);
-	struct st33zp24_platform_data *pin_infos =
-		((struct i2c_client *)TPM_VPRIV(chip))->dev.platform_data;
+	struct tpm_chip *chip = dev_get_drvdata(dev);
+	struct st33zp24_platform_data *pin_infos = dev->platform_data;
 	int ret = 0;
 
drivers/char/tpm/tpm_i2c_stm_st33.c +8 -11
 	if (power_mgt)
@@ -839,7 +837,7 @@ static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
 	else{
 		if (chip->data_buffer == NULL)
 			chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-		ret = tpm_pm_suspend(&client->dev);
+		ret = tpm_pm_suspend(dev);
 	}
 	return ret;
 }				/* tpm_st33_i2c_suspend() */
@@ -849,12 +847,10 @@ static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
  * @param: client, the i2c_client drescription (TPM I2C description).
  * @return: 0 in case of success.
  */
-static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
+static int tpm_st33_i2c_pm_resume(struct device *dev)
 {
-	struct tpm_chip *chip =
-	    (struct tpm_chip *)i2c_get_clientdata(client);
-	struct st33zp24_platform_data *pin_infos =
-		((struct i2c_client *)TPM_VPRIV(chip))->dev.platform_data;
+	struct tpm_chip *chip = dev_get_drvdata(dev);
+	struct st33zp24_platform_data *pin_infos = dev->platform_data;
 
 	int ret = 0;
 
@@ -867,7 +863,7 @@ static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
 	} else{
 	if (chip->data_buffer == NULL)
 		chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-	ret = tpm_pm_resume(&client->dev);
+	ret = tpm_pm_resume(dev);
 	if (!ret)
 		tpm_do_selftest(chip);
 	}
@@ -880,16 +876,15 @@ static const struct i2c_device_id 
 tpm_st33_i2c_id[] = {  };
 
 MODULE_DEVICE_TABLE(i2c, tpm_st33_i2c_id);
-
+static SIMPLE_DEV_PM_OPS(tpm_st33_i2c_ops, tpm_st33_i2c_pm_suspend, 
+tpm_st33_i2c_pm_resume);
 static struct i2c_driver tpm_st33_i2c_driver = {
 	.driver = {
 		   .owner = THIS_MODULE,
 		   .name = TPM_ST33_I2C,
+		   .pm = &tpm_st33_i2c_ops,
 		   },
 	.probe = tpm_st33_i2c_probe,
 	.remove = tpm_st33_i2c_remove,
-	.resume = tpm_st33_i2c_pm_resume,
-	.suspend = tpm_st33_i2c_pm_suspend,
 	.id_table = tpm_st33_i2c_id
 };
 
--

1.7.8.6

#29 lore raw

Re: [PATCH] char/tpm: Use struct dev_pm_ops for power management

In reply to Peter Huewe
Quoted 4 lines · Peter Huewe

This patch converts the suspend and resume functions for tpm_i2c_stm_st33 to the new dev_pm_ops.

One minor tweak, the PM funcs need to be inside CONFIG_PM_SLEEP to
avoid warnings when compiled without PM support. Applied with that change only.

Thanks Peter!
Kent

Quoted 88 lines

---
Note:
I'm sending this patch on behalf of myself only and nobody else,
especially not my employer - and I'm doing this in my spare time.
I have to inform you that I'm working for a direct competitor of the driver author, but I'm not using any knowledge that is considered secret in any way.
I have a personal interest in the TPM subsystem and want to keep it as clean as possible.

 drivers/char/tpm/tpm_i2c_stm_st33.c |   25 ++++++++++---------------
 1 files changed, 10 insertions(+), 15 deletions(-)
drivers/char/tpm/tpm_i2c_stm_st33.c +10 -15
index b11201f..ecbb81e 100644
@@ -826,12 +826,10 @@ static __devexit int tpm_st33_i2c_remove(struct i2c_client *client)
  * @param: mesg, the power management message.
  * @return: 0 in case of success.
  */
-static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
+static int tpm_st33_i2c_pm_suspend(struct device *dev)
 {
-	struct tpm_chip *chip =
-	    (struct tpm_chip *)i2c_get_clientdata(client);
-	struct st33zp24_platform_data *pin_infos =
-		((struct i2c_client *)TPM_VPRIV(chip))->dev.platform_data;
+	struct tpm_chip *chip = dev_get_drvdata(dev);
+	struct st33zp24_platform_data *pin_infos = dev->platform_data;
 	int ret = 0;
 
 	if (power_mgt)
@@ -839,7 +837,7 @@ static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
 	else{
 		if (chip->data_buffer == NULL)
 			chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-		ret = tpm_pm_suspend(&client->dev);
+		ret = tpm_pm_suspend(dev);
 	}
 	return ret;
 }				/* tpm_st33_i2c_suspend() */
@@ -849,12 +847,10 @@ static int tpm_st33_i2c_pm_suspend(struct i2c_client *client, pm_message_t mesg)
  * @param: client, the i2c_client drescription (TPM I2C description).
  * @return: 0 in case of success.
  */
-static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
+static int tpm_st33_i2c_pm_resume(struct device *dev)
 {
-	struct tpm_chip *chip =
-	    (struct tpm_chip *)i2c_get_clientdata(client);
-	struct st33zp24_platform_data *pin_infos =
-		((struct i2c_client *)TPM_VPRIV(chip))->dev.platform_data;
+	struct tpm_chip *chip = dev_get_drvdata(dev);
+	struct st33zp24_platform_data *pin_infos = dev->platform_data;
 
 	int ret = 0;
 
@@ -867,7 +863,7 @@ static int tpm_st33_i2c_pm_resume(struct i2c_client *client)
 	} else{
 	if (chip->data_buffer == NULL)
 		chip->data_buffer = pin_infos->tpm_i2c_buffer[0];
-	ret = tpm_pm_resume(&client->dev);
+	ret = tpm_pm_resume(dev);
 	if (!ret)
 		tpm_do_selftest(chip);
 	}
@@ -880,16 +876,15 @@ static const struct i2c_device_id tpm_st33_i2c_id[] = {
 };
 
 MODULE_DEVICE_TABLE(i2c, tpm_st33_i2c_id);
-
+static SIMPLE_DEV_PM_OPS(tpm_st33_i2c_ops, tpm_st33_i2c_pm_suspend, tpm_st33_i2c_pm_resume);
 static struct i2c_driver tpm_st33_i2c_driver = {
 	.driver = {
 		   .owner = THIS_MODULE,
 		   .name = TPM_ST33_I2C,
+		   .pm = &tpm_st33_i2c_ops,
 		   },
 	.probe = tpm_st33_i2c_probe,
 	.remove = tpm_st33_i2c_remove,
-	.resume = tpm_st33_i2c_pm_resume,
-	.suspend = tpm_st33_i2c_pm_suspend,
 	.id_table = tpm_st33_i2c_id
 };
 
Signature · 1 line
1.7.8.6
maintainer
#30 lore raw

Re: [PATCH] char/tpm: Use struct dev_pm_ops for power management

Am Donnerstag, 6. Dezember 2012, 17:27:02 schrieb Kent Yoder:

Quoted 12 lines · Peter Huewe
Quoted 4 lines

This patch converts the suspend and resume functions for tpm_i2c_stm_st33 to the new dev_pm_ops.

One minor tweak, the PM funcs need to be inside CONFIG_PM_SLEEP to
avoid warnings when compiled without PM support. Applied with that change only.

Thanks Peter!
Kent

Great to hear!

Peter