Re: [PATCH] Topcliff: Update PCH_SPI driver to 2.6.35

From: Grant Likely
Date: Sat Aug 14 2010 - 02:50:22 EST


2010/8/10 Masayuki Ohtak <masa-korg@xxxxxxxxxxxxxxx>:
> SPI driver of Topcliff PCH
>
> Topcliff PCH is the platform controller hub that is going to be used in
> Intel's upcoming general embedded platform. All IO peripherals in
> Topcliff PCH are actually devices sitting on AMBA bus.
> Topcliff PCH has SPI I/F. Using this I/F, it is able to access system
> devices connected to SPI.
>
> Signed-off-by: Masayuki Ohtake <masa-korg@xxxxxxxxxxxxxxx>

Hi Masayuki. Thanks for the patch. The driver is mostly structured
well and looks fairly good. However, there are quite a few stylistic
issues that should be easy to clean up, and a few technical concerns.
Comments below.

BTW, please make sure patches get sent out as plain-text, not html formatted.

> ---
> drivers/spi/Kconfig | 26 +
> drivers/spi/Makefile | 4 +
> drivers/spi/pch_spi.h | 177 +++
> drivers/spi/pch_spi_main.c | 1823
> ++++++++++++++++++++++++++++++++
> drivers/spi/pch_spi_platform_devices.c | 56 +
> 5 files changed, 2086 insertions(+), 0 deletions(-)
> create mode 100644 drivers/spi/pch_spi.h
> create mode 100644 drivers/spi/pch_spi_main.c
> create mode 100644 drivers/spi/pch_spi_platform_devices.c
>
> diff --git a/drivers/spi/Kconfig b/drivers/spi/Kconfig
> index f55eb01..b6ae72a 100644
> --- a/drivers/spi/Kconfig
> +++ b/drivers/spi/Kconfig
> @@ -53,6 +53,32 @@ if SPI_MASTER
>
> comment "SPI Master Controller Drivers"
>
> +config PCH_SPI_PLATFORM_DEVICE
> + boolean
> + default y if PCH_SPI
> + help
> + This driver is for PCH SPI of Topcliff which is an IOH for x86
> + embedded processor.

What is IOH and abbreviation for?

> + This registers SPI devices for a given board.
> +
> +config PCH_SPI_PLATFORM_DEVICE_COUNT
> + int "PCH SPI Bus count"
> + range 1 2
> + depends on PCH_SPI_PLATFORM_DEVICE
> + help
> + This driver is for PCH SPI of Topcliff which is an IOH for x86
> + embedded processor.
> + The number of SPI buses/channels supported by the PCH SPI controller.
> + PCH SPI of Topcliff supports only one channel.

Being required to specific this at kernel config time isn't
multiplatform friendly. Can the driver detect the number of busses at
runtime.

> +
> +config PCH_SPI
> + tristate "PCH SPI Controller"
> + depends on PCI
> + help
> + This driver is for PCH SPI of Topcliff which is an IOH for x86
> + embedded processor.
> + This driver can access PCH SPI bus device.

Put config PCH_SPI above PCH_SPI_PLATFORM_DEVICE and make
PCH_SPI_PLATFORM_DEVICE depend on PCH_SPI. (In fact, since
PCH_SPI_PLATFORM_DEVICE is always yes when PCH_SPI is selected, then
you probably don't need two configuration signals).

> +
> config SPI_ATMEL
> tristate "Atmel SPI Controller"
> depends on (ARCH_AT91 || AVR32)
> diff --git a/drivers/spi/Makefile b/drivers/spi/Makefile
> index f3d2810..aecc873 100644
> --- a/drivers/spi/Makefile
> +++ b/drivers/spi/Makefile
> @@ -59,3 +59,7 @@ obj-$(CONFIG_SPI_TLE62X0) += tle62x0.o
>
> # SPI slave drivers (protocol for that link)
> # ... add above this line ...
> +obj-$(CONFIG_PCH_SPI_PLATFORM_DEVICE) += pch_spi_platform_devices.o
> +obj-$(CONFIG_PCH_SPI) += pch_spi.o
> +pch_spi-objs := pch_spi_main.o

No need to use pch_spi-objs when there is only one .o file. Just name
the .c file what you want the resulting kernel module to be named.
Also, please rename the modules to "spi_pch_platform_device.o" and
"spi_pch.o" (I'm enforcing all new spi drivers to start with the
prefix "spi_").

Finally, do pch_spi_platform_devices.o and pch_spi_main.o really need
to be in separate modules?

> +
> diff --git a/drivers/spi/pch_spi.h b/drivers/spi/pch_spi.h
> new file mode 100644
> index 0000000..b40377a
> --- /dev/null
> +++ b/drivers/spi/pch_spi.h
> @@ -0,0 +1,177 @@
> +/*
> + * Copyright (C) 2010 OKI SEMICONDUCTOR Co., LTD.
> + *

It is helpful to have a one line description of what this file is for
in at the top of the header comment block.

[...]
> +#define PCH_SET_BITMSK(var, bitmask) ((var) |= (bitmask))
> +#define PCH_CLR_BITMSK(var, bitmask) ((var) &= (~(bitmask)))

Because the macro has side-effects, I'd prefer to see static inlines
used here with an all lowercase name.

> +
> +/**
> + * struct pch_spi_data - Holds the SPI channel specific details
> + * @io_remap_addr: The remapped PCI base address
> + * @p_master: Pointer to the SPI master structure
> + * @work: Reference to work queue handler
> + * @p_wk_q: Workqueue for carrying out execution of the
> + * requests
> + * @wait: Wait queue for waking up upon receiving an
> + * interrupt.
> + * @b_transfer_complete: Status of SPI Transfer
> + * @bcurrent_msg_processing: Status flag for message processing
> + * @lock: Lock for protecting this structure
> + * @queue: SPI Message queue
> + * @status: Status of the SPI driver
> + * @bpw_len: Length of data to be transferred in bits per
> + * word
> + * @b_transfer_active: Flag showing active transfer
> + * @tx_index: Transmit data count; for bookkeeping during
> + * transfer
> + * @rx_index: Receive data count; for bookkeeping during
> + * transfer
> + * @tx_buff: Buffer for data to be transmitted
> + * @rx_index: Buffer for Received data
> + * @n_curnt_chip: The chip number that this SPI driver currently
> + * operates on
> + * @p_current_chip: Reference to the current chip that this SPI
> + * driver currently operates on
> + * @p_current_msg: The current message that this SPI driver is
> + * handling
> + * @p_cur_trans: The current transfer that this SPI driver is
> + * handling
> + * @p_board_dat: Reference to the SPI device data structure
> + */
> +struct pch_spi_data {
> + void __iomem *io_remap_addr;
> + struct spi_master *p_master;
> + struct work_struct work;
> + struct workqueue_struct *p_wk_q;
> + wait_queue_head_t wait;
> + u8 b_transfer_complete;
> + u8 bcurrent_msg_processing;
> + spinlock_t lock;
> + struct list_head queue;
> + u8 status;
> + u32 bpw_len;
> + s8 b_transfer_active;
> + u32 tx_index;
> + u32 rx_index;
> + u16 *pkt_tx_buff;
> + u16 *pkt_rx_buff;
> + u8 n_curnt_chip;
> + struct spi_device *p_current_chip;
> + struct spi_message *p_current_msg;
> + struct spi_transfer *p_cur_trans;
> + struct pch_spi_board_data *p_board_dat;
> +};
> +
> +/**
> + * struct pch_spi_board_data - Holds the SPI device specific details
> + * @pdev: Pointer to the PCI device
> + * @irq_reg_sts: Status of IRQ registration
> + * @pci_req_sts: Status of pci_request_regions
> + * @suspend_sts: Status of suspend
> + * @p_data: Pointer to SPI channel data structure
> + */
> +struct pch_spi_board_data {
> + struct pci_dev *pdev;
> + u8 irq_reg_sts;
> + u8 pci_req_sts;
> + u8 suspend_sts;
> + struct pch_spi_data *p_data[PCH_MAX_DEV];
> +};
> +#endif
> diff --git a/drivers/spi/pch_spi_main.c b/drivers/spi/pch_spi_main.c
> new file mode 100644
> index 0000000..da87d92
> --- /dev/null
> +++ b/drivers/spi/pch_spi_main.c
> @@ -0,0 +1,1823 @@
> +/*
> + * Copyright (C) 2010 OKI SEMICONDUCTOR Co., LTD.
> + *
> + * 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; version 2 of the License.
> + *
> + * 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., 59 Temple Place, Suite 330, Boston, MA 02111-1307,
> USA.
> + */
> +
> +#include <linux/pci.h>
> +#include <linux/wait.h>
> +#include <linux/spi/spi.h>
> +#include <linux/interrupt.h>
> +#include <linux/sched.h>
> +#include <linux/spi/spidev.h>
> +#include <linux/module.h>
> +#include <linux/device.h>
> +#include <linux/mutex.h>
> +#include "pch_spi.h"
> +
> +static struct pci_device_id pch_spi_pcidev_id[] = {
> + {PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_GE_SPI)},
> + {0,}
> +};
> +
> +static void (*pch_spi_gcbptr) (struct pch_spi_data *);
> +static DEFINE_MUTEX(pch_spi_mutex);

Are these statics really necessary? I think the callback can be
removed (see my comment below). The mutex should probably be a member
of the pch_spi_data structure. Getting rid of the static symbols will
make it easier if the driver ever needs to support multiple instances
of the device.

> +/**
> + * ch_spi_disable_interrupts() - Disables specified interrupts
> + * @master: Pointer to struct spi_master.
> + * @interrupt: Interrups to be enabled.
> + */
> +static void pch_spi_disable_interrupts(struct spi_master *master, u8
> interrupt)
> +{
> + u32 reg_val_spcr;
> +
> + reg_val_spcr = pch_spi_readreg(master, PCH_SPCR);
> +
> + dev_dbg(&master->dev, "%s SPCR content =%x\n",
> + __func__, reg_val_spcr);
> +
> + switch (interrupt & 0x1f) {

Is it possible for more than one of these interrupt bits to be set at
the same time? If so, then this switch() statement won't always work.

> + case PCH_RFI:
> + /*clear RFIE bit in SPCR */
> + dev_dbg(&master->dev,
> + "%s clearing RFI in pch_spi_disable_interrupts\n", __func__);
> + PCH_CLR_BITMSK(reg_val_spcr, SPCR_RFIE_BIT);
> + break;
> +
> + case PCH_TFI:
> + /*clear TFIE bit in SPCR */
> + dev_dbg(&master->dev,
> + "%s clearing TFI in pch_spi_disable_interrupts\n", __func__);
> + PCH_CLR_BITMSK(reg_val_spcr, SPCR_TFIE_BIT);
> + break;
> +
> + case PCH_FI:
> + /*clear FIE bit in SPCR */
> + dev_dbg(&master->dev,
> + "%s clearing FI in pch_spi_disable_interrupts\n", __func__);
> + PCH_CLR_BITMSK(reg_val_spcr, SPCR_FIE_BIT);
> + break;
> +
> + case PCH_ORI:
> + /*clear ORIE bit in SPCR */
> + dev_dbg(&master->dev,
> + "%s clearing ORI in pch_spi_disable_interrupts\n", __func__);
> + PCH_CLR_BITMSK(reg_val_spcr, SPCR_ORIE_BIT);
> + break;
> +
> + case PCH_MDFI:
> + /*clear MODFIE bit in SPCR */
> + dev_dbg(&master->dev,
> + "%s clearing MDFI in pch_spi_disable_interrupts\n", __func__);
> + PCH_CLR_BITMSK(reg_val_spcr, SPCR_MDFIE_BIT);
> + break;
> +
> + default:
> + dev_info(&master->dev,
> + "%s Unknown interrupt(=0x%02x)\n", __func__, interrupt & 0x1f);
> + }
> +
> + pch_spi_writereg(master, PCH_SPCR, reg_val_spcr);
> +
> + dev_dbg(&master->dev, "%s SPCR after disabling interrupts =%x\n",
> + __func__, reg_val_spcr);
> +}
> +
> +/**
> + * pch_spi_handler() - Interrupt handler
> + * @irq: The interrupt number.
> + * @dev_id: Pointer to struct pch_spi_board_data.
> + */
> +static irqreturn_t pch_spi_handler(int irq, void *dev_id)
> +{
> + /*channel & read/write indices */
> + int dev, read_cnt;
> +
> + /*SPSR content */
> + u32 reg_spsr_val, reg_spcr_val;
> +
> + /*book keeping variables */
> + u32 n_read, tx_index, rx_index, bpw_len;
> +
> + /*to hold channel data */
> + struct pch_spi_data *p_data;
> +
> + /*buffer to store rx/tx data */
> + u16 *pkt_rx_buffer, *pkt_tx_buff;
> +
> + /*register addresses */
> + void __iomem *spsr;
> + void __iomem *spdrr;
> + void __iomem *spdwr;
> +
> + /*remapped pci base address */
> + void __iomem *io_remap_addr;
> +
> + irqreturn_t tRetVal = IRQ_NONE;

nit: use lowercase for variable names.

> +
> + struct pch_spi_board_data *p_board_dat =
> + (struct pch_spi_board_data *)dev_id;
> +
> + if (p_board_dat->suspend_sts == true) {
> + dev_dbg(&p_board_dat->pdev->dev,
> + "%s returning due to suspend\n", __func__);
> + } else {
> + for (dev = 0; dev < PCH_MAX_DEV; dev++) {
> + p_data = p_board_dat->p_data[dev];
> + io_remap_addr = p_data->io_remap_addr;
> + spsr = io_remap_addr + PCH_SPSR;
> +
> + reg_spsr_val = ioread32(spsr);
> +
> + /*Check if the interrupt is for SPI device */
> +
> + if (reg_spsr_val & (SPSR_FI_BIT | SPSR_RFI_BIT)) {
> + dev_dbg(&p_board_dat->pdev->dev,
> + "SPSR in %s=%x\n", __func__, reg_spsr_val);
> + /*clear interrupt */
> + iowrite32(reg_spsr_val, spsr);
> +
> + if (p_data->b_transfer_active == true) {
> + rx_index = p_data->rx_index;
> + tx_index = p_data->tx_index;
> + bpw_len = p_data->bpw_len;
> + pkt_rx_buffer = p_data->pkt_rx_buff;
> + pkt_tx_buff = p_data->pkt_tx_buff;
> +
> + spdrr = io_remap_addr + PCH_SPDRR;
> + spdwr = io_remap_addr + PCH_SPDWR;
> +
> + n_read = PCH_READABLE(reg_spsr_val);
> +
> + for (read_cnt = 0; (read_cnt < n_read);
> + read_cnt++) {
> + /*read data */
> + pkt_rx_buffer[rx_index++] =
> + ioread32(spdrr);
> + /*write data */
> +
> + if (tx_index < bpw_len) {
> + iowrite32
> + (pkt_tx_buff
> + [tx_index++],
> + spdwr);
> + }
> + }
> +
> + /*disable RFI if not needed */
> + if ((bpw_len - rx_index) <=
> + PCH_MAX_FIFO_DEPTH) {
> + dev_dbg(&p_board_dat->pdev->dev,
> + "%s disabling RFI as data "
> + "remaining=%d\n", __func__,
> + (bpw_len - rx_index));
> +
> + reg_spcr_val =
> + ioread32(io_remap_addr +
> + PCH_SPCR);
> +
> + /*disable RFI */
> + PCH_CLR_BITMSK(reg_spcr_val,
> + SPCR_RFIE_BIT);
> +
> + /*reset rx threshold */
> + reg_spcr_val &=
> + MASK_RFIC_SPCR_BITS;
> + reg_spcr_val |=
> + (PCH_RX_THOLD_MAX <<
> + SPCR_RFIC_FIELD);
> +
> + iowrite32(PCH_CLR_BITMSK
> + (reg_spcr_val,
> + SPCR_RFIE_BIT),
> + (io_remap_addr +
> + PCH_SPCR));
> + }
> +
> + /*update counts */
> + p_data->tx_index = tx_index;
> +
> + p_data->rx_index = rx_index;
> +
> + dev_dbg(&p_board_dat->pdev->dev,
> + "%s rx_index=%d tx_index=%d "
> + "nWritable=%d n_read=%d\n",
> + __func__, rx_index, tx_index,
> + (16 -
> + (PCH_WRITABLE(reg_spsr_val))),
> + n_read);
> +
> + }
> +
> + /*if transfer complete interrupt */
> + if (reg_spsr_val & SPSR_FI_BIT) {
> + dev_dbg(&p_board_dat->pdev->dev,
> + "%s FI bit in SPSR"
> + " set\n", __func__);
> +
> + /*disable FI & RFI interrupts */
> + pch_spi_disable_interrupts(p_data->
> + p_master, PCH_FI | PCH_RFI);
> +
> + /*transfer is completed;inform
> + pch_spi_process_messages */
> +
> + if (pch_spi_gcbptr != NULL) {
> + dev_dbg(&p_board_dat->pdev->dev,
> + "%s invoking callback\n",
> + __func__);
> + (*pch_spi_gcbptr) (p_data);
> + }
> + }
> +
> + tRetVal = IRQ_HANDLED;
> + }
> + }
> + }

Wow, the nesting on this function is very deep, and hard to follow.
Candidate for refactoring?

> +
> + dev_dbg(&p_board_dat->pdev->dev, "%s EXIT return value=%d\n",
> + __func__, tRetVal);
> +
> + return tRetVal;
> +}
> +
> +/**
> + * pch_spi_entcb() - Registers the callback function
> + * @pch_spi_cb: Callback function pointer.
> + */
> +static void pch_spi_entcb(void (*pch_spi_cb) (struct pch_spi_data *))
> +{
> + if (pch_spi_cb != NULL)
> + pch_spi_gcbptr = pch_spi_cb;
> +}

This trivial function is used exactly once in the probe routine. Just
set the variable in-place. In fact, this doesn't need to be a
callback at all because it is *always* set to pch_spi_callback.
Removing it will make the driver simpler.

> +/**
> + * pch_spi_setup_transfe() - Configures the PCH SPI hardware for transfer
> + * @spi: Pointer to struct spi_device.
> + */
> +static void pch_spi_setup_transfer(struct spi_device *spi)
> +{
> + u32 reg_spcr_val;
> +
> + dev_dbg(&spi->dev, "%s SPBRR content =%x setting baud rate=%d\n",
> + __func__, pch_spi_readreg(spi->master, PCH_SPBRR), spi->max_speed_hz);
> +
> + pch_spi_set_baud_rate(spi->master, spi->max_speed_hz);
> +
> + /*set bits per word */
> + dev_dbg(&spi->dev, "%s :setting bits_per_word=%d\n",
> + __func__, spi->bits_per_word);
> + pch_spi_set_bits_per_word(spi->master, spi->bits_per_word);
> +
> + dev_dbg(&spi->dev,
> + "%s SPBRR content after setting baud rate & bits per word=%x\n",
> + __func__, pch_spi_readreg(spi->master, PCH_SPBRR));
> +
> + reg_spcr_val = pch_spi_readreg(spi->master, PCH_SPCR);
> + dev_dbg(&spi->dev, "%s SPCR content = %x\n", __func__, reg_spcr_val);
> +
> + /*set bit justification */
> +
> + if ((spi->mode & SPI_LSB_FIRST) != 0) {
> + /*LSB first */
> + PCH_CLR_BITMSK(reg_spcr_val, SPCR_LSBF_BIT);
> + dev_dbg(&spi->dev, "%s :setting LSBF bit to 0\n", __func__);
> + } else {
> + /*MSB first */
> + PCH_SET_BITMSK(reg_spcr_val, SPCR_LSBF_BIT);
> + dev_dbg(&spi->dev, "%s :setting LSBF bit to 1\n", __func__);
> + }

There are a lot of these if/else blocks which call
PCH_{SET,CLR}_BITMSK(). The code could be simplified with a helper
function that implements the two paths. Something like
pch_spi_clrsetbits() perhaps.


> +static int pch_spi_transfer(struct spi_device *pspi, struct spi_message
> *pmsg)
> +{
> +
> + struct spi_transfer *p_transfer;
> + struct pch_spi_data *p_data = spi_master_get_devdata(pspi->master);
> + int retval;
> + int ret;
> +
> + ret = mutex_lock_interruptible(&pch_spi_mutex);
> + if (ret) {
> + retval = -ERESTARTSYS;
> + goto err_return;
> + }
> +
> + /*validate spi message and baud rate */
> + if (unlikely((list_empty(&pmsg->transfers) == 1) ||
> + ((pspi->max_speed_hz) == 0))) {
> + if (list_empty(&pmsg->transfers) == 1)
> + dev_err(&pspi->dev, "%s list empty\n", __func__);
> +
> + if ((pspi->max_speed_hz) == 0) {
> + dev_err(&pspi->dev, "%s pch_spi_tranfer maxspeed=%d\n",
> + __func__, (pspi->max_speed_hz));
> + }
> + dev_err(&pspi->dev, "%s returning EINVAL\n", __func__);
> +
> + retval = -EINVAL;
> + goto err_return_mutex;
> + }
> +
> + dev_dbg(&pspi->dev,
> + "%s Transfer List not empty. Transfer Speed is set.\n", __func__);
> +
> + /*validate Tx/Rx buffers and Transfer length */
> + list_for_each_entry(p_transfer, &pmsg->transfers,
> + transfer_list) {
> + if ((((p_transfer->tx_buf) == NULL)
> + && ((p_transfer->rx_buf) == NULL))
> + || (p_transfer->len == 0)) {
> + if (((p_transfer->tx_buf) == NULL)
> + && ((p_transfer->rx_buf) == NULL)) {
> + dev_err(&pspi->dev,
> + "%s Tx and Rx buffer NULL\n", __func__);
> + }
> +
> + if (p_transfer->len == 0) {
> + dev_err(&pspi->dev, "%s Transfer"
> + " length invalid\n", __func__);
> + }
> +
> + dev_err(&pspi->dev, "%s returning EINVAL\n", __func__);
> +
> + retval = -EINVAL;
> + goto err_return_mutex;
> + }
> +
> + dev_dbg(&pspi->dev,
> + "%s Tx/Rx buffer valid. Transfer length valid\n", __func__);
> +
> + /*if baud rate hs been specified validate the same */
> + if (p_transfer->speed_hz) {
> + if ((p_transfer->speed_hz) >
> + PCH_MAX_BAUDRATE) {
> + retval = -EINVAL;
> + dev_err(&pspi->dev,
> + "%s Invalid Baud rate\n", __func__);
> + goto err_return_mutex;
> + }

If the requested speed is too fast, then the maximum bus speed should
be used instead.

> + /*copy Tx Data */
> + if ((p_data->p_cur_trans->tx_buf) != NULL) {
> + for (j = 0; j < (p_data->bpw_len); j++) {
> + if (PCH_8_BPW == *bpw) {
> + p_data->pkt_tx_buff[j] =
> + (((u8 *) (p_data->p_cur_trans->
> + tx_buf))[j]);
> + dev_dbg(&p_data->p_master->dev, "%s=%x\n",
> + __func__, (p_data->pkt_tx_buff[j]));
> + } else {
> + p_data->pkt_tx_buff[j] =
> + ((u16 *) (p_data->p_cur_trans->
> + tx_buf))[j];
> + dev_dbg(&p_data->p_master->dev,
> + "%s xmt data = %x\n", __func__,
> + (p_data->pkt_tx_buff[j]));
> + }
> + }

Optimization note; by coding it this way, the (PCH_8_BPW == *bpw) test
is being performed on every iteration through the loop. If you had
separate iterators for 8 and 16 bit transfer, the code will be faster.

> + }
> +
> + /*if len greater than PCH_MAX_FIFO_DEPTH,
> + write 16,else len bytes */

Nit: It would be easier to read if the comment blocks were tidied up.
There should be a space between /* and the first word, and some
comments, like this one, will fit on a single line.

> + if ((p_data->bpw_len) > PCH_MAX_FIFO_DEPTH)
> + n_writes = PCH_MAX_FIFO_DEPTH;
> + else
> + n_writes = (p_data->bpw_len);
> +
> + dev_dbg(&p_data->p_master->dev,
> + "\n%s:Pulling down SSN low - writing 0x2 to SSNXCR\n", __func__);

When splitting lines in the middle of an argument list, please indent
up to the same level as the first argument.

> +static void pch_spi_copy_rx_data(struct pch_spi_data *p_data, int bpw)
> +{
> + int j;
> + /*copy Rx Data */
> + if ((p_data->p_cur_trans->rx_buf) != NULL) {

if (p_data->pcur_trans->rx_buf)
return;

That reduces the indentation on the rest of the function by one level
and makes it easier to read and understand.

> + for (j = 0; j < (p_data->bpw_len); j++) {
> + if (PCH_8_BPW == bpw) {
> + ((u8 *) (p_data->p_cur_trans->rx_buf))[j] =
> + (u8) ((p_data->pkt_rx_buff[j]) & 0xFF);
> +
> + dev_dbg(&p_data->p_master->dev,
> + "rcv data in %s =%x\n", __func__,
> + (p_data->pkt_rx_buff[j]));
> +
> + } else {
> + ((u16 *) (p_data->p_cur_trans->rx_buf))[j] =
> + (u16) (p_data->pkt_rx_buff[j]);
> + dev_dbg(&p_data->p_master->dev,
> + "rcv data in %s = %x\n", __func__,
> + (p_data->pkt_rx_buff[j]));
> + }
> + }

Same comment on optimization.

> + }
> +}
> +
> +
> +static void pch_spi_process_messages(struct work_struct *pwork)
> +{
> + struct spi_message *pmsg;
> + int bpw;
> +
> + struct pch_spi_data *p_data =
> + container_of(pwork, struct pch_spi_data, work);
> + dev_dbg(&p_data->p_master->dev,
> + "%s p_data initialized\n", __func__);
> +
> + spin_lock(&p_data->lock);
> +
> + /*check if suspend has been initiated;if yes flush queue */
> + if ((p_data->p_board_dat->suspend_sts)
> + || (p_data->status == STATUS_EXITING)) {
> + dev_dbg(&p_data->p_master->dev,
> + "%s suspend/remove initiated,flushing queue\n", __func__);
> + list_for_each_entry(pmsg, p_data->queue.next, queue) {
> + pmsg->status = -EIO;
> +
> + if (pmsg->complete != 0)
> + pmsg->complete(pmsg->context);

The complete callback cannot be called while still holding the spinlock.

> +
> + /*delete from queue */
> + list_del_init(&pmsg->queue);
> + }
> +
> + spin_unlock(&p_data->lock);

Put a return statement here. It will eliminate the else { }
indentation below; again improving readability.

> + } else {
> + p_data->bcurrent_msg_processing = true;
> + dev_dbg(&p_data->p_master->dev,
> + "%s Set p_data->bcurrent_msg_processing= true\n",
> + __func__);
> +
> + /*Get the message from the queue and delete it from there. */
> + p_data->p_current_msg =
> + list_entry(p_data->queue.next, struct spi_message,
> + queue);
> + dev_dbg(&p_data->p_master->dev,
> + "%s :Got new message from queue\n", __func__);
> + list_del_init(&p_data->p_current_msg->queue);
> +
> + p_data->p_current_msg->status = 0;
> +
> + dev_dbg(&p_data->p_master->dev,
> + "%s :Invoking pch_spi_select_chip\n", __func__);
> + pch_spi_select_chip(p_data, p_data->p_current_msg->spi);
> +
> + spin_unlock(&p_data->lock);
> +
> + do {
> + /*If we are already processing a message get the next
> + transfer structure from the message otherwise retrieve
> + the 1st transfer request from the message. */
> + spin_lock(&p_data->lock);
> +
> + if (p_data->p_cur_trans == NULL) {
> + p_data->p_cur_trans =
> + list_entry(p_data->p_current_msg->transfers.
> + next, struct spi_transfer,
> + transfer_list);
> + dev_dbg(&p_data->p_master->dev,
> + "%s :Getting 1st transfer message\n",
> + __func__);
> + } else {
> + p_data->p_cur_trans =
> + list_entry(p_data->p_cur_trans->
> + transfer_list.next,
> + struct spi_transfer,
> + transfer_list);
> + dev_dbg(&p_data->p_master->dev,
> + "%s :Getting next transfer message\n",
> + __func__);
> + }
> +
> + spin_unlock(&p_data->lock);
> +
> + pch_spi_set_tx(p_data, &bpw, &pmsg);
> +
> + /*Control interrupt*/
> + pch_spi_set_ir(p_data);
> +
> + /*Disable SPI transfer */
> + dev_dbg(&p_data->p_master->dev,
> + "%s:invoking pch_spi_set_enable to "
> + "disable spi transfer\n", __func__);
> + pch_spi_set_enable((p_data->p_current_chip), false);
> +
> + /*clear FIFO */
> + dev_dbg(&p_data->p_master->dev,
> + "%s:invoking pch_spi_clear_fifo to clear fifo\n",
> + __func__);
> + pch_spi_clear_fifo(p_data->p_master);
> +
> + /*copy Rx Data */
> + pch_spi_copy_rx_data(p_data, bpw);
> +
> + /*free memory */
> + kfree(p_data->pkt_rx_buff);
> + if (p_data->pkt_rx_buff)
> + p_data->pkt_rx_buff = NULL;
> +
> + kfree(p_data->pkt_tx_buff);
> + if (p_data->pkt_tx_buff)
> + p_data->pkt_tx_buff = NULL;
> +
> + /*increment message count */
> + p_data->p_current_msg->actual_length +=
> + p_data->p_cur_trans->len;
> +
> + dev_dbg(&p_data->p_master->dev,
> + "%s:p_data->p_current_msg->actual_length=%d\n",
> + __func__, p_data->p_current_msg->actual_length);
> +
> + /*check for delay */
> + if (p_data->p_cur_trans->delay_usecs) {
> + dev_dbg
> + (&p_data->p_master->dev, "%s:"
> + "delay in usec=%d\n", __func__,
> + p_data->p_cur_trans->delay_usecs);
> + udelay(p_data->p_cur_trans->delay_usecs);
> + }
> +
> + spin_lock(&p_data->lock);
> +
> + /*No more transfer in this message. */
> + if ((p_data->p_cur_trans->transfer_list.next) ==
> + &(p_data->p_current_msg->transfers)) {
> + pch_spi_nomore_transfer(p_data, pmsg);
> + }
> +
> + spin_unlock(&p_data->lock);
> +
> + } while ((p_data->p_cur_trans) != NULL);
> + }
> +}
> +
> +static int pch_spi_probe(struct pci_dev *pdev, const struct pci_device_id
> *id)
> +{
> +
> + struct spi_master *p_master[PCH_MAX_DEV];
> +
> + struct pch_spi_board_data *p_board_dat;
> + int retval, i, j, k, l, m;

nit: for loop index variables should be reused. Don't use a new index
for every single for() loop.

> + int spi_alloc_num, spi_master_num;
> +
> + dev_dbg(&pdev->dev, "%s ENTRY\n", __func__);
> + /* initialize the call back function */
> + pch_spi_entcb(pch_spi_callback);
> + dev_dbg(&pdev->dev, "%s invoked pch_spi_entcb\n", __func__);
> +
> + /* allocate memory for private data */
> + p_board_dat = kmalloc(sizeof(struct pch_spi_board_data), GFP_KERNEL);

kzalloc()

> +
> + if (p_board_dat == NULL) {
> + dev_err(&pdev->dev,
> + " %s memory allocation for private data failed\n",
> + __func__);
> + retval = -ENOMEM;
> + goto err_kmalloc;
> + }
> +
> + dev_dbg(&pdev->dev,
> + "%s memory allocation for private data success\n", __func__);
> +
> + /* enable PCI device */
> + retval = pci_enable_device(pdev);
> + if (retval != 0) {
> + dev_err(&pdev->dev, "%s pci_enable_device FAILED\n", __func__);
> +
> + goto err_pci_en_device;
> + }
> +
> + dev_dbg(&pdev->dev, "%s pci_enable_device returned=%d\n",
> + __func__, retval);
> +
> + p_board_dat->pdev = pdev;
> +
> + /* alllocate memory for SPI master */
> + for (i = 0, spi_alloc_num = 0; i < PCH_MAX_DEV; i++, spi_alloc_num++) {
> + p_master[i] = spi_alloc_master(&pdev->dev,
> + sizeof(struct pch_spi_data));
> +
> + if (p_master[i] == NULL) {
> + retval = -ENOMEM;
> + dev_err(&pdev->dev, "%s [ch=%d]Fail.\n", __func__, i);
> + goto err_spi_alloc_master;
> + }
> + }
> +
> + dev_dbg(&pdev->dev,
> + "%s spi_alloc_master returned non NULL\n", __func__);
> +
> + /* initialize members of SPI master */
> + for (j = 0, spi_master_num = 0; j < PCH_MAX_DEV;
> + j++, spi_master_num++) {

spi_master_num isn't used except for the error path, and it always has
the same value as 'j'. Get it out of the loop initializer! :-) It
can be set in the error path.

> + p_master[j]->bus_num = 0;

You probably want -1 here so that the spi layer dynamically assigns an
spi bus number.

> + p_master[j]->num_chipselect = PCH_MAX_CS;
> + p_master[j]->setup = pch_spi_setup;
> + dev_dbg(&pdev->dev,
> + "%s setup member of SPI master initialized\n",
> + __func__);
> + p_master[j]->transfer = pch_spi_transfer;
> + dev_dbg(&pdev->dev,
> + "%s transfer member of SPI master initialized\n", __func__);
> + p_master[j]->cleanup = pch_spi_cleanup;
> + dev_dbg(&pdev->dev,
> + "%s cleanup member of SPI master initialized\n", __func__);
> +
> + p_board_dat->p_data[j] =
> + spi_master_get_devdata(p_master[j]);

Unnecessary line break. There are a lot of these in this driver.

> +
> + p_board_dat->p_data[j]->p_master = p_master[j];
> + p_board_dat->p_data[j]->n_curnt_chip = 255;
> + p_board_dat->p_data[j]->p_current_chip = NULL;
> + p_board_dat->p_data[j]->b_transfer_complete = false;
> + p_board_dat->p_data[j]->pkt_tx_buff = NULL;
> + p_board_dat->p_data[j]->pkt_rx_buff = NULL;
> + p_board_dat->p_data[j]->tx_index = 0;
> + p_board_dat->p_data[j]->rx_index = 0;
> + p_board_dat->p_data[j]->b_transfer_active = false;

The above 7 lines can be dropped when the code is changed to use
kzalloc(). kzalloc() clears the allocated memory.

> + p_board_dat->p_data[j]->p_board_dat = p_board_dat;
> + p_board_dat->suspend_sts = false;
> +
> + /* Register the controller with the SPI core. */
> + retval = spi_register_master(p_master[j]);
> + if (retval != 0) {
> + dev_err(&pdev->dev,
> + "%s spi_register_master FAILED\n", __func__);
> + goto err_spi_reg_master;
> + }
> +
> + dev_dbg(&pdev->dev, "%s spi_register_master returned=%d\n",
> + __func__, retval);
> + }
> +
> + /* allocate resources for PCH SPI */
> + retval = pch_spi_get_resources(p_board_dat);
> + if (retval != 0) {
> + dev_err(&pdev->dev, "%s fail(retval=%d)\n",
> + __func__, retval);
> + goto err_spi_get_resources;
> + }
> +
> + dev_dbg(&pdev->dev, "%s pch_spi_get_resources returned=%d\n",
> + __func__, retval);
> +
> + /* save private data in dev */
> + pci_set_drvdata(pdev, (void *)p_board_dat);

Remove the (void *) cast. It should not be necessary.

> + dev_dbg(&pdev->dev, "%s invoked pci_set_drvdata\n", __func__);
> +
> + /* set master mode */
> + for (k = 0; k < PCH_MAX_DEV; k++) {
> + pch_spi_set_master_mode(p_master[k]);
> + dev_dbg(&pdev->dev,
> + "%s invoked pch_spi_set_master_mode\n", __func__);
> + }
> +
> + return 0;
> +
> +err_spi_get_resources:
> + for (l = 0; l <= spi_master_num; l++)
> + spi_unregister_master(p_master[l]);
> +err_spi_reg_master:
> +err_spi_alloc_master:
> + for (m = 0; m <= spi_alloc_num; m++)
> + spi_master_put(p_master[m]);
> + pci_disable_device(pdev);
> +err_pci_en_device:
> + kfree(p_board_dat);
> +err_kmalloc:
> + return retval;
> +}
> +
> +static void pch_spi_remove(struct pci_dev *pdev)
> +{
> + int i;
> +
> + struct pch_spi_board_data *p_board_dat = pci_get_drvdata(pdev);
> +
> + dev_dbg(&pdev->dev, "%s ENTRY\n", __func__);
> +
> + if (p_board_dat != NULL) {

if (!p_board_dat) {
dev_err(...);
return;
}

> + dev_dbg(&pdev->dev, "%s invoked pci_get_drvdata\n", __func__);
> +
> + /* check for any pending messages */
> + if ((-EBUSY) == pch_spi_check_request_pending(p_board_dat)) {
> + dev_dbg(&pdev->dev,
> + "%s pch_spi_check_request_pending returned EBUSY\n",
> + __func__);
> + /* no need to take any particular action;
> + proceed with remove even
> + though queue is not empty */
> + }
> +
> + dev_dbg(&pdev->dev,
> + "%s pch_spi_check_request_pending invoked\n", __func__);
> +
> + /* Free resources allocated for PCH SPI */
> + pch_spi_free_resources(p_board_dat);
> + dev_dbg(&pdev->dev,
> + "%s invoked pch_spi_free_resources\n", __func__);
> +
> + /* Unregister SPI master */
> + for (i = 0; i < PCH_MAX_DEV; i++) {
> + spi_unregister_master(p_board_dat->p_data[i]->
> + p_master);
> + dev_dbg(&pdev->dev,
> + "%s invoked spi_unregister_master\n", __func__);
> + }
> +
> + /* free memory for private data */
> + kfree(p_board_dat);
> +
> + pci_set_drvdata(pdev, NULL);
> +
> + dev_dbg(&pdev->dev,
> + "%s memory for private data freed\n", __func__);
> +
> + /* disable PCI device */
> + pci_disable_device(pdev);
> +
> + dev_dbg(&pdev->dev,
> + "%s invoked pci_disable_device\n", __func__);
> +
> + } else {
> + dev_err(&pdev->dev,
> + "%s pci_get_drvdata returned NULL\n", __func__);
> + }
> +}
> +
> +#ifdef CONFIG_PM
> +static int pch_spi_suspend(struct pci_dev *pdev, pm_message_t state)
> +{
> + int i;
> + u8 count;
> + s32 retval;
> +
> + struct pch_spi_board_data *p_board_dat = pci_get_drvdata(pdev);
> +
> + dev_dbg(&pdev->dev, "%s ENTRY\n", __func__);
> +
> + if (p_board_dat == NULL) {
> + dev_err(&pdev->dev,
> + "%s pci_get_drvdata returned NULL\n", __func__);
> + retval = -EFAULT;

Ditto here; return -EFAULT;

> + } else {
> + retval = 0;
> + dev_dbg(&pdev->dev,
> + "%s pci_get_drvdata invoked successfully\n", __func__);
> + p_board_dat->suspend_sts = true;
> + dev_dbg(&pdev->dev,
> + "%s p_board_dat->bSuspending set to true\n", __func__);
> +
> + /* check if the current message is processed:
> + Only after thats done the transfer will be suspended */
> + for (i = 0; i < PCH_MAX_DEV; i++) {
> + count = 255;
> +
> + while ((--count) > 0) {
> + if (p_board_dat->p_data[i]->
> + bcurrent_msg_processing == false) {
> + dev_dbg(&pdev->dev, "%s p_board_dat->"
> + "p_data->bCurrent_"
> + "msg_processing = false\n",
> + __func__);
> + break;
> + } else {
> + dev_dbg(&pdev->dev, "%s p_board_dat->"
> + "p_data->bCurrent_msg_"
> + "processing = true\n",
> + __func__);
> + }
> + msleep(PCH_SLEEP_TIME);
> + }
> + }
> +
> + /* Free IRQ */
> + if (p_board_dat->irq_reg_sts == true) {
> + /* disable all interrupts */
> + for (i = 0; i < PCH_MAX_DEV; i++) {
> + pch_spi_disable_interrupts(p_board_dat->
> + p_data[i]->
> + p_master,
> + PCH_ALL);
> + pch_spi_reset(p_board_dat->p_data[i]->
> + p_master);
> + dev_dbg(&pdev->dev,
> + "%s pch_spi_disable_interrupts invoked"
> + "successfully\n", __func__);
> + }
> +
> + free_irq(p_board_dat->pdev->irq, (void *)p_board_dat);
> +
> + p_board_dat->irq_reg_sts = false;
> + dev_dbg(&pdev->dev,
> + "%s free_irq invoked successfully."
> + "p_data->irq_reg_sts = false\n",
> + __func__);
> + }
> +
> + /* save config space */
> + retval = pci_save_state(pdev);
> +
> + if (retval == 0) {
> + dev_dbg(&pdev->dev, "%s pci_save_state returned=%d\n",
> + __func__, retval);
> + /* disable PM notifications */
> + pci_enable_wake(pdev, PCI_D3hot, 0);
> + dev_dbg(&pdev->dev,
> + "%s pci_enable_wake invoked successfully\n",
> + __func__);
> + /* disable PCI device */
> + pci_disable_device(pdev);
> + dev_dbg(&pdev->dev,
> + "%s pci_disable_device invoked successfully\n",
> + __func__);
> + /* move device to D3hot state */
> + pci_set_power_state(pdev, PCI_D3hot);
> + dev_dbg(&pdev->dev,
> + "%s pci_set_power_state invoked successfully\n",
> + __func__);
> + } else {
> + dev_err(&pdev->dev,
> + "%s pci_save_state failed\n", __func__);
> + }
> + }
> +
> + dev_dbg(&pdev->dev, "%s return=%d\n", __func__, retval);
> +
> + return retval;
> +}
> +
> +static int pch_spi_resume(struct pci_dev *pdev)
> +{
> + int i;
> + s32 retval;
> +
> + struct pch_spi_board_data *p_board_dat = pci_get_drvdata(pdev);
> + dev_dbg(&pdev->dev, "%s ENTRY\n", __func__);
> +
> + if (p_board_dat == NULL) {
> + dev_err(&pdev->dev,
> + "%s pci_get_drvdata returned NULL\n", __func__);
> + retval = -EFAULT;
> + } else {
> + retval = 0;
> + /* move device to DO power state */
> + pci_set_power_state(pdev, PCI_D0);
> + dev_dbg(&pdev->dev,
> + "%s pci_set_power_state invoked successfully\n", __func__);
> +
> + /* restore state */
> + pci_restore_state(pdev);
> + dev_dbg(&pdev->dev,
> + "%s pci_restore_state invoked successfully\n", __func__);
> +
> + retval = pci_enable_device(pdev);
> + if (retval < 0) {
> + dev_err(&pdev->dev,
> + "%s pci_enable_device failed\n", __func__);
> + } else {
> + dev_dbg(&pdev->dev,
> + "%s pci_enable_device returned=%d\n",
> + __func__, retval);
> +
> + /* disable PM notifications */
> + pci_enable_wake(pdev, PCI_D3hot, 0);
> + dev_dbg(&pdev->dev,
> + "%s pci_enable_wake invoked successfully\n", __func__);
> +
> + /* register IRQ handler */
> + if ((p_board_dat->irq_reg_sts) != true) {
> + /* register IRQ */
> + retval = request_irq(p_board_dat->pdev->irq,
> + pch_spi_handler, IRQF_SHARED,
> + DRIVER_NAME,
> + p_board_dat);
> + if (retval < 0) {
> + dev_err(&pdev->dev,
> + "%s request_irq failed\n", __func__);
> + } else {
> + dev_dbg(&pdev->dev, "%s request_irq"
> + " returned=%d\n", __func__, retval);
> + p_board_dat->irq_reg_sts = true;
> +
> + /* reset PCH SPI h/w */
> + for (i = 0; i < PCH_MAX_DEV; i++) {
> + pch_spi_reset(p_board_dat->
> + p_data[i]->
> + p_master);
> + dev_dbg(&pdev->dev,
> + "pdev->dev, %s pch_spi_reset "
> + "invoked successfully\n",
> + __func__);
> + pch_spi_set_master_mode
> + (p_board_dat->p_data[i]->
> + p_master);
> + dev_dbg(&pdev->dev,
> + "%s pch_spi_set_master_mode "
> + "invoked successfully\n",
> + __func__);
> + }
> +
> + /* set suspend status to false */
> + p_board_dat->suspend_sts = false;
> +
> + dev_dbg(&pdev->dev, "%s set p_board_dat"
> + "->bSuspending = false\n", __func__);
> + }
> + }
> + }
> + }

Another very deeply indented function.

> +
> + dev_dbg(&pdev->dev, "%s returning=%d\n", __func__, retval);
> +
> + return retval;
> +}
> +#else
> +#define pch_spi_suspend NULL
> +#define pch_spi_resume NULL
> +
> +#endif
> +
> +static struct pci_driver pch_spi_pcidev = {
> + .name = "pch_spi",
> + .id_table = pch_spi_pcidev_id,
> + .probe = pch_spi_probe,
> + .remove = pch_spi_remove,
> + .suspend = pch_spi_suspend,
> + .resume = pch_spi_resume,
> +};
> +
> +static int __init pch_spi_init(void)
> +{
> + return pci_register_driver(&pch_spi_pcidev);
> +}
> +
> +
> +static void __exit pch_spi_exit(void)
> +{
> + pci_unregister_driver(&pch_spi_pcidev);
> +}
> +
> +MODULE_LICENSE("GPL");
> +MODULE_DESCRIPTION("PCH SPI PCI Driver");
> +module_init(pch_spi_init);
> +module_exit(pch_spi_exit);
> +
> diff --git a/drivers/spi/pch_spi_platform_devices.c
> b/drivers/spi/pch_spi_platform_devices.c
> new file mode 100644
> index 0000000..a0d6488
> --- /dev/null
> +++ b/drivers/spi/pch_spi_platform_devices.c
> @@ -0,0 +1,56 @@
> +/*
> + * Copyright (C) 2010 OKI SEMICONDUCTOR Co., LTD.
> + *
> + * 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; version 2 of the License.
> + *
> + * 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., 59 Temple Place, Suite 330, Boston, MA 02111-1307,
> USA.
> + */
> +
> +#include <linux/module.h>
> +#include <linux/spi/spidev.h>
> +#include <linux/device.h>
> +#include <linux/spi/spi.h>
> +
> +static struct spi_board_info pch_spi_slaves[] = {
> + {
> + .modalias = "spidev",
> + .max_speed_hz = 1000000,
> + .bus_num = 0,
> + .chip_select = 0,
> + .platform_data = NULL,
> + .mode = SPI_MODE_0,
> + },
> +#if (CONFIG_PCH_SPI_PLATFORM_DEVICE_COUNT - 1)
> + {
> + .modalias = "spidev",
> + .max_speed_hz = 1000000,
> + .bus_num = 1,
> + .chip_select = 1,
> + .platform_data = NULL,
> + .mode = SPI_MODE_0,
> + },
> +#endif
> +};

This file is misnamed; these aren't platform_devices, they are spi_devices.

> +
> +static __init int Load(void)
> +{
> + int iRetVal = -1;

int rc = -1;

> +
> + if (!spi_register_board_info
> + (pch_spi_slaves, ARRAY_SIZE(pch_spi_slaves)))
> + iRetVal = 0;
> +
> + return iRetVal;
> +}
> +
> +module_init(Load);

How about when the module is removed?

This particular module (not the rest of the driver though) is really
rather a hack. It seems like you need a method for creating spidev
devices from userspace at runtime, or having a method for specifying
spi device from the kernel command line. Using a flattened device
tree fragment would work as well.

g.


--
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/