Skip to contentSkip to conversation

Add modules for realtek USB card reader

PATCH Superseded [email protected] 4 participants 14 messages Jan 14, 2014 – Jan 20, 2014

Subsystems (3)

v2 of 7 Superseded Latest is v7 →

lore
Share & tools
lore
#0 lore raw

v2 cover Add modules for realtek USB card reader

From: Roger Tseng <[email protected]>

This patchset adds modules to support Realtek USB vendor specific class flash card reader: one base module in MFD subsystem and two host modules in both mmc and memstick subsystems. The architecture is similar to rtsx_pci.

This work is done primarily to replace the staging driver: staging/rts5139, which doesn't utilize mmc nor memstick subsystems. Once the patchset or its revision is applied, we may need Greg's help to remove the staging one.

v2:
1. Fix potential buffer overflow in rtsx_usb_seq_write_register()
2. Remove unnecessary casting and messages
2. Revise coding style and white spaces

Roger Tseng (3):
mfd: Add realtek USB card reader driver
mmc: Add realtek USB sdmmc host driver
memstick: Add realtek USB memstick host driver

 drivers/memstick/host/Kconfig       |   10 +
 drivers/memstick/host/Makefile      |    1 +
 drivers/memstick/host/rtsx_usb_ms.c |  855 ++++++++++++++++++++
 drivers/mfd/Kconfig                 |   10 +
 drivers/mfd/Makefile                |    1 +
 drivers/mfd/rtsx_usb.c              |  788 ++++++++++++++++++
 drivers/mmc/host/Kconfig            |    7 +
 drivers/mmc/host/Makefile           |    1 +
 drivers/mmc/host/rtsx_usb_sdmmc.c   | 1509 +++++++++++++++++++++++++++++++++++
 include/linux/mfd/rtsx_usb.h        |  619 ++++++++++++++
 10 files changed, 3801 insertions(+)
 create mode 100644 drivers/memstick/host/rtsx_usb_ms.c
 create mode 100644 drivers/mfd/rtsx_usb.c
 create mode 100644 drivers/mmc/host/rtsx_usb_sdmmc.c
 create mode 100644 include/linux/mfd/rtsx_usb.h
Signature · 1 line
1.8.2
#1 lore raw

PATCH v2 1/3 mfd: Add realtek USB card reader driver

From: Roger Tseng <[email protected]>

Realtek USB card reader provides a channel to transfer command or data to flash memory cards. This driver exports host instances for mmc and memstick subsystems and handles basic works.

---
 drivers/mfd/Kconfig          |  10 +
 drivers/mfd/Makefile         |   1 +
 drivers/mfd/rtsx_usb.c       | 788 +++++++++++++++++++++++++++++++++++++++++++
 include/linux/mfd/rtsx_usb.h | 619 +++++++++++++++++++++++++++++++++
 4 files changed, 1418 insertions(+)
 create mode 100644 drivers/mfd/rtsx_usb.c
 create mode 100644 include/linux/mfd/rtsx_usb.h
drivers/mfd/Kconfig +10 -0
index b7c74a7..4c99ebd 100644
@@ -507,6 +507,16 @@ config MFD_RTSX_PCI
 	  types of memory cards, such as Memory Stick, Memory Stick Pro,
 	  Secure Digital and MultiMediaCard.
 
+config MFD_RTSX_USB
+	tristate "Realtek USB card reader"
+	depends on USB
+	select MFD_CORE
+	help
+	Select this option to get support for Realtek USB 2.0 card readers
+	including RTS5129, RTS5139, RTS5179 and RTS5170.
+	Realtek card reader supports access to many types of memory cards,
+	such as Memory Stick Pro, Secure Digital and MultiMediaCard.
+
 config MFD_RC5T583
 	bool "Ricoh RC5T583 Power Management system device"
 	depends on I2C=y
drivers/mfd/Makefile +1 -0
index 8a28dc9..33b8de6 100644
@@ -14,6 +14,7 @@ obj-$(CONFIG_MFD_CROS_EC_SPI)	+= cros_ec_spi.o
 
 rtsx_pci-objs			:= rtsx_pcr.o rts5209.o rts5229.o rtl8411.o rts5227.o rts5249.o
 obj-$(CONFIG_MFD_RTSX_PCI)	+= rtsx_pci.o
+obj-$(CONFIG_MFD_RTSX_USB)	+= rtsx_usb.o
 
 obj-$(CONFIG_HTC_EGPIO)		+= htc-egpio.o
 obj-$(CONFIG_HTC_PASIC3)	+= htc-pasic3.o
drivers/mfd/rtsx_usb.c new file +788 -0

788 lines not on this page · raw message

include/linux/mfd/rtsx_usb.h new file +619 -0

619 lines not on this page · raw message

Signature · 1 line
1.8.2
#2 lore raw

Re: [PATCH 1/3] mfd: Add realtek USB card reader driver

Quoted 24 lines

From: Roger Tseng <[email protected]>

Realtek USB card reader provides a channel to transfer command or data to flash memory cards. This driver exports host instances for mmc and memstick subsystems and handles basic works.

---
 drivers/mfd/Kconfig          |  10 +
 drivers/mfd/Makefile         |   1 +
 drivers/mfd/rtsx_usb.c       | 788 +++++++++++++++++++++++++++++++++++++++++++
 include/linux/mfd/rtsx_usb.h | 619 +++++++++++++++++++++++++++++++++
 4 files changed, 1418 insertions(+)
 create mode 100644 drivers/mfd/rtsx_usb.c
 create mode 100644 include/linux/mfd/rtsx_usb.h
drivers/mfd/Kconfig +0 -0
index b7c74a7..4c99ebd 100644
@@ -507,6 +507,16 @@ config MFD_RTSX_PCI
 	  types of memory cards, such as Memory Stick, Memory Stick Pro,
 	  Secure Digital and MultiMediaCard.
 

drivers/mfd/Kconfig config MFD_RTSX_PCI · line 519

Quoted 10 lines
drivers/mfd/Kconfig +10 -0
+config MFD_RTSX_USB
+	tristate "Realtek USB card reader"
+	depends on USB
+	select MFD_CORE
+	help
+	Select this option to get support for Realtek USB 2.0 card readers
+	including RTS5129, RTS5139, RTS5179 and RTS5170.
+	Realtek card reader supports access to many types of memory cards,
+	such as Memory Stick Pro, Secure Digital and MultiMediaCard.
+

The help section should be indented by 2 spaces.

Quoted 43 lines
drivers/mfd/Kconfig +0 -0
 config MFD_RC5T583
 	bool "Ricoh RC5T583 Power Management system device"
 	depends on I2C=y
drivers/mfd/Makefile +1 -0
index 8a28dc9..33b8de6 100644
@@ -14,6 +14,7 @@ obj-$(CONFIG_MFD_CROS_EC_SPI)	+= cros_ec_spi.o
 
 rtsx_pci-objs			:= rtsx_pcr.o rts5209.o rts5229.o rtl8411.o rts5227.o rts5249.o
 obj-$(CONFIG_MFD_RTSX_PCI)	+= rtsx_pci.o
+obj-$(CONFIG_MFD_RTSX_USB)	+= rtsx_usb.o
 
 obj-$(CONFIG_HTC_EGPIO)		+= htc-egpio.o
 obj-$(CONFIG_HTC_PASIC3)	+= htc-pasic3.o
drivers/mfd/rtsx_usb.c new file +22 -0
new file mode 100644
index 0000000..905ec8d
@@ -0,0 +1,788 @@
+/* Driver for Realtek USB card reader
+ *
+ * Copyright(c) 2009-2013 Realtek Semiconductor Corp. All rights reserved.
+ *
+ * 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, 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, see <http://www.gnu.org/licenses/>.
+ *
+ * Author:
+ *   Roger Tseng <[email protected]>
+ */
+#include <linux/module.h>
+#include <linux/slab.h>

drivers/mfd/rtsx_usb.c line 32

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+#include <linux/mutex.h>
+#include <linux/usb.h>
+#include <linux/platform_device.h>
+#include <linux/mfd/core.h>
+#include <asm/unaligned.h>
+#include <linux/mfd/rtsx_usb.h>
+
+static int polling_pipe = 1;
+module_param(polling_pipe, int, S_IRUGO | S_IWUSR);
+MODULE_PARM_DESC(polling_pipe, "polling pipe (0: ctl, 1: bulk)");

'/n' here.

drivers/mfd/rtsx_usb.c line 40

Quoted 8 lines
drivers/mfd/rtsx_usb.c new file +8 -0
+static struct mfd_cell rtsx_usb_cells[] = {
+	[RTSX_USB_SD_CARD] = {
+		.name = DRV_NAME_RTSX_USB_SDMMC,
+	},
+	[RTSX_USB_MS_CARD] = {
+		.name = DRV_NAME_RTSX_USB_MS,
+	},
+};

I'm not entirely convinced that this is an MFD device?

drivers/mfd/rtsx_usb.c line 44

Quoted 3 lines
drivers/mfd/rtsx_usb.c new file +3 -0
+static void rtsx_usb_sg_timed_out(unsigned long data)
+{
+	struct rtsx_ucr *ucr = (struct rtsx_ucr *)data;

What's going to happen when your device runs 64bit?

drivers/mfd/rtsx_usb.c line 47

Quoted 2 lines
drivers/mfd/rtsx_usb.c new file +2 -0
+	dev_dbg(&ucr->pusb_intf->dev, "%s: sg transfer timed out", __func__);
+	usb_sg_cancel(&ucr->current_sg);

Are you sure this needs to live here?

Why isn't this in drivers/usb?

Quoted 17 lines
drivers/mfd/rtsx_usb.c new file +17 -0
+	/* we know the cancellation is caused by time-out */
+	ucr->current_sg.status = -ETIMEDOUT;
+}
+
+static int rtsx_usb_bulk_transfer_sglist(struct rtsx_ucr *ucr,
+		unsigned int pipe, struct scatterlist *sg, int num_sg,
+		unsigned int length, unsigned int *act_len, int timeout)
+{
+	int ret;
+
+	dev_dbg(&ucr->pusb_intf->dev, "%s: xfer %u bytes, %d entries\n",
+			__func__, length, num_sg);
+	ret = usb_sg_init(&ucr->current_sg, ucr->pusb_dev, pipe, 0,
+			sg, num_sg, length, GFP_NOIO);
+	if (ret)
+		return ret;
+

drivers/mfd/rtsx_usb.c line 75

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+	ucr->sg_timer.expires = jiffies + msecs_to_jiffies(timeout);
+	add_timer(&ucr->sg_timer);
+	usb_sg_wait(&ucr->current_sg);
+	del_timer(&ucr->sg_timer);
+
+	if (act_len)
+		*act_len = ucr->current_sg.bytes;
+
+	return ucr->current_sg.status;
+}

Code looks fine, but shouldn't this live an a USB driver?

Quoted 11 lines
drivers/mfd/rtsx_usb.c new file +11 -0
+int rtsx_usb_transfer_data(struct rtsx_ucr *ucr, unsigned int pipe,
+			      void *buf, unsigned int len, int num_sg,
+			      unsigned int *act_len, int timeout)
+{
+	if (timeout < 600)
+		timeout = 600;
+
+	if (num_sg)
+		return rtsx_usb_bulk_transfer_sglist(ucr, pipe,
+				(struct scatterlist *)buf, num_sg, len, act_len,
+				timeout);

drivers/mfd/rtsx_usb.c line 97

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+	else
+		return usb_bulk_msg(ucr->pusb_dev, pipe, buf, len, act_len,
+				timeout);
+}
+EXPORT_SYMBOL_GPL(rtsx_usb_transfer_data);
+
+static int rtsx_usb_seq_write_register(struct rtsx_ucr *ucr,
+		u16 addr, u16 len, u8 *data)
+{
+	u16 cmd_len = len + 12;

Why 12? Please #define.

drivers/mfd/rtsx_usb.c line 94

Quoted 1 line
drivers/mfd/rtsx_usb.c new file +1 -0
+	if (data == NULL)

if (!data)

drivers/mfd/rtsx_usb.c line 106

Quoted 7 lines
drivers/mfd/rtsx_usb.c new file +7 -0
+		return -EINVAL;
+
+	if (cmd_len > IOBUF_SIZE)
+		return -EINVAL;
+
+	if (cmd_len % 4)
+		cmd_len += (4 - cmd_len % 4);

Please document in a comment.

drivers/mfd/rtsx_usb.c line 103

Quoted 2 lines
drivers/mfd/rtsx_usb.c new file +2 -0
+
+

Extra '/n'

drivers/mfd/rtsx_usb.c line 118

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+	ucr->cmd_buf[0] = 'R';
+	ucr->cmd_buf[1] = 'T';
+	ucr->cmd_buf[2] = 'C';
+	ucr->cmd_buf[3] = 'R';
+	ucr->cmd_buf[PACKET_TYPE] = SEQ_WRITE;
+	ucr->cmd_buf[5] = (u8)(len >> 8);
+	ucr->cmd_buf[6] = (u8)len;
+	ucr->cmd_buf[STAGE_FLAG] = 0;
+	ucr->cmd_buf[8] = (u8)(addr >> 8);
+	ucr->cmd_buf[9] = (u8)addr;

I think someone said there are kernel macros for this stuff.

drivers/mfd/rtsx_usb.c line 124

Quoted 5 lines
drivers/mfd/rtsx_usb.c new file +5 -0
+	memcpy(ucr->cmd_buf + 12, data, len);
+
+	return rtsx_usb_transfer_data(ucr,
+			usb_sndbulkpipe(ucr->pusb_dev, EP_BULK_OUT),
+			ucr->cmd_buf, cmd_len, 0, NULL, 100);

Still think it should be a USB driver!

drivers/mfd/rtsx_usb.c line 133

Quoted 9 lines
drivers/mfd/rtsx_usb.c new file +9 -0
+}
+
+static int rtsx_usb_seq_read_register(struct rtsx_ucr *ucr,
+		u16 addr, u16 len, u8 *data)
+{
+	int i, ret;
+	u16 rsp_len, res_len;
+
+	if (data == NULL)

if (!data)

Quoted 7 lines
drivers/mfd/rtsx_usb.c new file +7 -0
+		return -EINVAL;
+
+	res_len = len % 4;
+	rsp_len = len - res_len;
+
+	/* 4-byte aligned part */
+	if (rsp_len) {

drivers/mfd/rtsx_usb.c line 150

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+		ucr->cmd_buf[0] = 'R';
+		ucr->cmd_buf[1] = 'T';
+		ucr->cmd_buf[2] = 'C';
+		ucr->cmd_buf[3] = 'R';
+		ucr->cmd_buf[PACKET_TYPE] = SEQ_READ;
+		ucr->cmd_buf[5] = (u8)(rsp_len >> 8);
+		ucr->cmd_buf[6] = (u8)rsp_len;
+		ucr->cmd_buf[STAGE_FLAG] = STAGE_R;
+		ucr->cmd_buf[8] = (u8)(addr >> 8);
+		ucr->cmd_buf[9] = (u8)addr;

This looks the same as above. If so, please place in a function.

Quoted 13 lines
drivers/mfd/rtsx_usb.c new file +13 -0
+		ret = rtsx_usb_transfer_data(ucr,
+				usb_sndbulkpipe(ucr->pusb_dev, EP_BULK_OUT),
+				ucr->cmd_buf, 12, 0, NULL, 100);
+		if (ret)
+			return ret;
+
+		ret = rtsx_usb_transfer_data(ucr,
+				usb_rcvbulkpipe(ucr->pusb_dev, EP_BULK_IN),
+				data, rsp_len, 0, NULL, 100);
+		if (ret)
+			return ret;
+	}
+

drivers/mfd/rtsx_usb.c line 174

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+	/* unaligned part */
+	for (i = 0; i < res_len; i++) {
+		ret = rtsx_usb_read_register(ucr, addr + rsp_len + i,
+				data + rsp_len + i);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}

'/n' here.

Quoted 12 lines
drivers/mfd/rtsx_usb.c new file +12 -0
+int rtsx_usb_read_ppbuf(struct rtsx_ucr *ucr, u8 *buf, int buf_len)
+{
+	return rtsx_usb_seq_read_register(ucr, PPBUF_BASE2, (u16)buf_len, buf);
+}
+EXPORT_SYMBOL_GPL(rtsx_usb_read_ppbuf);
+
+int rtsx_usb_write_ppbuf(struct rtsx_ucr *ucr, u8 *buf, int buf_len)
+{
+	return rtsx_usb_seq_write_register(ucr, PPBUF_BASE2, (u16)buf_len, buf);
+}
+EXPORT_SYMBOL_GPL(rtsx_usb_write_ppbuf);
+

drivers/mfd/rtsx_usb.c line 196

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+int rtsx_usb_ep0_write_register(struct rtsx_ucr *ucr, u16 addr,
+		u8 mask, u8 data)
+{
+	u16 value = 0, index = 0;
+
+	value |= 0x03 << 14;
+	value |= addr & 0x3FFF;
+	value = ((value << 8) & 0xFF00) | ((value >> 8) & 0x00FF);
+	index |= (u16)mask;
+	index |= (u16)data << 8;

Lots of random numbers here, please #define for clarity and ease of reading.

drivers/mfd/rtsx_usb.c line 200

Quoted 3 lines
drivers/mfd/rtsx_usb.c new file +3 -0
+	return usb_control_msg(ucr->pusb_dev,
+			usb_sndctrlpipe(ucr->pusb_dev, 0), 0, 0x40,
+			value, index, NULL, 0, 100);

Same here.

drivers/mfd/rtsx_usb.c line 208

Quoted 8 lines
drivers/mfd/rtsx_usb.c new file +8 -0
+}
+EXPORT_SYMBOL_GPL(rtsx_usb_ep0_write_register);
+
+int rtsx_usb_ep0_read_register(struct rtsx_ucr *ucr, u16 addr, u8 *data)
+{
+	u16 value = 0;
+
+	if (data == NULL)

if (!data), etc

drivers/mfd/rtsx_usb.c line 214

Quoted 6 lines
drivers/mfd/rtsx_usb.c new file +6 -0
+		return -EINVAL;
+	*data = 0;
+
+	value |= 0x02 << 14;
+	value |= addr & 0x3FFF;
+	value = ((value << 8) & 0xFF00) | ((value >> 8) & 0x00FF);

Less random numbers for #defines please, etc.

drivers/mfd/rtsx_usb.c line 226

Quoted 11 lines
drivers/mfd/rtsx_usb.c new file +11 -0
+	return usb_control_msg(ucr->pusb_dev,
+			usb_rcvctrlpipe(ucr->pusb_dev, 0), 0, 0xC0,
+			value, 0, data, 1, 100);
+}
+EXPORT_SYMBOL_GPL(rtsx_usb_ep0_read_register);
+
+void rtsx_usb_add_cmd(struct rtsx_ucr *ucr,
+		    u8 cmd_type,
+		    u16 reg_addr,
+		    u8 mask,
+		    u8 data)

This is a strange way of organising your function params.

drivers/mfd/rtsx_usb.c line 230

Quoted 4 lines
drivers/mfd/rtsx_usb.c new file +4 -0
+{
+	int i;
+
+	if (ucr->cmd_idx < ((IOBUF_SIZE - CMD_OFFSET) / 4)) {

I think you can drop a layer of ()'s here.

Quoted 31 lines
drivers/mfd/rtsx_usb.c new file +31 -0
+		i = CMD_OFFSET + ucr->cmd_idx * 4;
+
+		ucr->cmd_buf[i++] = ((cmd_type & 0x03) << 6) |
+			(u8)((reg_addr >> 8) & 0x3F);
+		ucr->cmd_buf[i++] = (u8)reg_addr;
+		ucr->cmd_buf[i++] = mask;
+		ucr->cmd_buf[i++] = data;
+
+		ucr->cmd_idx++;
+	}
+}
+EXPORT_SYMBOL_GPL(rtsx_usb_add_cmd);
+
+int rtsx_usb_send_cmd(struct rtsx_ucr *ucr, u8 flag, int timeout)
+{
+	int ret;
+
+	ucr->cmd_buf[CNT_H] = (u8)(ucr->cmd_idx >> 8);
+	ucr->cmd_buf[CNT_L] = (u8)(ucr->cmd_idx);
+	ucr->cmd_buf[STAGE_FLAG] = flag;
+
+	ret = rtsx_usb_transfer_data(ucr,
+			usb_sndbulkpipe(ucr->pusb_dev, EP_BULK_OUT),
+			ucr->cmd_buf, ucr->cmd_idx * 4 + CMD_OFFSET,
+			0, NULL, timeout);
+	if (ret) {
+		rtsx_usb_clear_fsm_err(ucr);
+		return ret;
+	}
+
+	return 0;

drivers/mfd/rtsx_usb.c line 271

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+}
+EXPORT_SYMBOL_GPL(rtsx_usb_send_cmd);
+
+int rtsx_usb_get_rsp(struct rtsx_ucr *ucr, int rsp_len, int timeout)
+{
+	if (rsp_len <= 0)
+		return -EINVAL;
+
+	if (rsp_len % 4)
+		rsp_len += (4 - rsp_len % 4);

Please document.

Quoted 95 lines
drivers/mfd/rtsx_usb.c new file +95 -0
+	return rtsx_usb_transfer_data(ucr,
+			usb_rcvbulkpipe(ucr->pusb_dev, EP_BULK_IN),
+			ucr->rsp_buf, rsp_len, 0, NULL, timeout);
+}
+EXPORT_SYMBOL_GPL(rtsx_usb_get_rsp);
+
+static int rtsx_usb_get_status_with_bulk(struct rtsx_ucr *ucr, u16 *status)
+{
+	int ret;
+
+	rtsx_usb_init_cmd(ucr);
+	rtsx_usb_add_cmd(ucr, READ_REG_CMD, CARD_EXIST, 0x00, 0x00);
+	rtsx_usb_add_cmd(ucr, READ_REG_CMD, OCPSTAT, 0x00, 0x00);
+	ret = rtsx_usb_send_cmd(ucr, MODE_CR, 100);
+	if (ret)
+		return ret;
+
+	ret = rtsx_usb_get_rsp(ucr, 2, 100);
+	if (ret)
+		return ret;
+
+	*status = ((ucr->rsp_buf[0] >> 2) & 0x0f) |
+		  ((ucr->rsp_buf[1] & 0x03) << 4);
+
+	return 0;
+}
+
+int rtsx_usb_get_card_status(struct rtsx_ucr *ucr, u16 *status)
+{
+	int ret;
+
+	if (status == NULL)
+		return -EINVAL;
+
+	if (polling_pipe == 0)
+		ret = usb_control_msg(ucr->pusb_dev,
+				usb_rcvctrlpipe(ucr->pusb_dev, 0),
+				0x02, 0xC0, 0, 0, status, 2, 100);
+	else
+		ret = rtsx_usb_get_status_with_bulk(ucr, status);
+
+	/* usb_control_msg may return positive when success */
+	if (ret < 0)
+		return ret;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(rtsx_usb_get_card_status);
+
+static int rtsx_usb_write_phy_register(struct rtsx_ucr *ucr, u8 addr, u8 val)
+{
+	dev_dbg(&ucr->pusb_intf->dev, "Write 0x%x to phy register 0x%x\n",
+			val, addr);
+
+	rtsx_usb_init_cmd(ucr);
+
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, HS_VSTAIN, 0xFF, val);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, HS_VCONTROL, 0xFF, addr & 0x0F);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, HS_VLOADM, 0xFF, 0x00);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, HS_VLOADM, 0xFF, 0x00);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, HS_VLOADM, 0xFF, 0x01);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, HS_VCONTROL,
+			0xFF, (addr >> 4) & 0x0F);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, HS_VLOADM, 0xFF, 0x00);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, HS_VLOADM, 0xFF, 0x00);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, HS_VLOADM, 0xFF, 0x01);
+
+	return rtsx_usb_send_cmd(ucr, MODE_C, 100);
+}
+
+int rtsx_usb_write_register(struct rtsx_ucr *ucr, u16 addr, u8 mask, u8 data)
+{
+	rtsx_usb_init_cmd(ucr);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, addr, mask, data);
+	return rtsx_usb_send_cmd(ucr, MODE_C, 100);
+}
+EXPORT_SYMBOL_GPL(rtsx_usb_write_register);
+
+int rtsx_usb_read_register(struct rtsx_ucr *ucr, u16 addr, u8 *data)
+{
+	int ret;
+
+	if (data != NULL)
+		*data = 0;
+
+	rtsx_usb_init_cmd(ucr);
+	rtsx_usb_add_cmd(ucr, READ_REG_CMD, addr, 0, 0);
+	ret = rtsx_usb_send_cmd(ucr, MODE_CR, 100);
+	if (ret)
+		return ret;
+
+	ret = rtsx_usb_get_rsp(ucr, 1, 100);
+	if (ret)
+		return ret;
+

drivers/mfd/rtsx_usb.c line 377

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+	if (data != NULL)
+		*data = ucr->rsp_buf[0];
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(rtsx_usb_read_register);
+
+static inline u8 double_ssc_depth(u8 depth)
+{
+	return ((depth > 1) ? (depth - 1) : depth);

You can drop a layer of ()'s here too.

drivers/mfd/rtsx_usb.c line 383

Quoted 6 lines
drivers/mfd/rtsx_usb.c new file +6 -0
+}
+
+static u8 revise_ssc_depth(u8 ssc_depth, u8 div)
+{
+	if (div > CLK_DIV_1) {
+		if (ssc_depth > (div - 1))

And here, etc.

Quoted 184 lines
drivers/mfd/rtsx_usb.c new file +184 -0
+			ssc_depth -= (div - 1);
+		else
+			ssc_depth = SSC_DEPTH_2M;
+	}
+
+	return ssc_depth;
+}
+
+int rtsx_usb_switch_clock(struct rtsx_ucr *ucr, unsigned int card_clock,
+		u8 ssc_depth, bool initial_mode, bool double_clk, bool vpclk)
+{
+	int ret;
+	u8 n, clk_divider, mcu_cnt, div;
+
+	if (!card_clock) {
+		ucr->cur_clk = 0;
+		return 0;
+	}
+
+	if (initial_mode) {
+		/* We use 250k(around) here, in initial stage */
+		clk_divider = SD_CLK_DIVIDE_128;
+		card_clock = 30000000;
+	} else {
+		clk_divider = SD_CLK_DIVIDE_0;
+	}
+
+	ret = rtsx_usb_write_register(ucr, SD_CFG1,
+			SD_CLK_DIVIDE_MASK, clk_divider);
+	if (ret < 0)
+		return ret;
+
+	card_clock /= 1000000;
+	dev_dbg(&ucr->pusb_intf->dev,
+			"Switch card clock to %dMHz\n", card_clock);
+
+	if (!initial_mode && double_clk)
+		card_clock *= 2;
+	dev_dbg(&ucr->pusb_intf->dev,
+			"Internal SSC clock: %dMHz (cur_clk = %d)\n",
+			card_clock, ucr->cur_clk);
+
+	if (card_clock == ucr->cur_clk)
+		return 0;
+
+	/* Converting clock value into internal settings: n and div */
+	n = card_clock - 2;
+	if ((card_clock <= 2) || (n > MAX_DIV_N))
+		return -EINVAL;
+
+	mcu_cnt = 60/card_clock + 3;
+	if (mcu_cnt > 15)
+		mcu_cnt = 15;
+
+	/* Make sure that the SSC clock div_n is not less than MIN_DIV_N */
+
+	div = CLK_DIV_1;
+	while (n < MIN_DIV_N && div < CLK_DIV_4) {
+		n = (n + 2) * 2 - 2;
+		div++;
+	}
+	dev_dbg(&ucr->pusb_intf->dev, "n = %d, div = %d\n", n, div);
+
+	if (double_clk)
+		ssc_depth = double_ssc_depth(ssc_depth);
+
+	ssc_depth = revise_ssc_depth(ssc_depth, div);
+	dev_dbg(&ucr->pusb_intf->dev, "ssc_depth = %d\n", ssc_depth);
+
+	rtsx_usb_init_cmd(ucr);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CLK_DIV, CLK_CHANGE, CLK_CHANGE);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CLK_DIV,
+			0x3F, (div << 4) | mcu_cnt);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, SSC_CTL1, SSC_RSTB, 0);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, SSC_CTL2,
+			SSC_DEPTH_MASK, ssc_depth);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, SSC_DIV_N_0, 0xFF, n);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, SSC_CTL1, SSC_RSTB, SSC_RSTB);
+	if (vpclk) {
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, SD_VPCLK0_CTL,
+				PHASE_NOT_RESET, 0);
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, SD_VPCLK0_CTL,
+				PHASE_NOT_RESET, PHASE_NOT_RESET);
+	}
+
+	ret = rtsx_usb_send_cmd(ucr, MODE_C, 2000);
+	if (ret < 0)
+		return ret;
+
+	ret = rtsx_usb_write_register(ucr, SSC_CTL1, 0xff,
+			SSC_RSTB | SSC_8X_EN | SSC_SEL_4M);
+	if (ret < 0)
+		return ret;
+
+	/* Wait SSC clock stable */
+	usleep_range(100, 1000);
+
+	ret = rtsx_usb_write_register(ucr, CLK_DIV, CLK_CHANGE, 0);
+	if (ret < 0)
+		return ret;
+
+	ucr->cur_clk = card_clock;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(rtsx_usb_switch_clock);
+
+int rtsx_usb_card_exclusive_check(struct rtsx_ucr *ucr, int card)
+{
+	int ret;
+	u16 val;
+	u16 cd_mask[] = {
+		[RTSX_USB_SD_CARD] = (CD_MASK & ~SD_CD),
+		[RTSX_USB_MS_CARD] = (CD_MASK & ~MS_CD)
+	};
+
+	ret = rtsx_usb_get_card_status(ucr, &val);
+	/*
+	 * If get status fails, return 0 (ok) for the exclusive check
+	 * and let the flow fail at somewhere else.
+	 */
+	if (ret)
+		return 0;
+
+	if (val & cd_mask[card])
+		return -EIO;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(rtsx_usb_card_exclusive_check);
+
+static int rtsx_usb_reset_chip(struct rtsx_ucr *ucr)
+{
+	int ret;
+	u8 val;
+
+	rtsx_usb_init_cmd(ucr);
+
+	if (CHECK_PKG(ucr, LQFP48)) {
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CARD_PWR_CTL,
+				LDO3318_PWR_MASK, LDO_SUSPEND);
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CARD_PWR_CTL,
+				FORCE_LDO_POWERB, FORCE_LDO_POWERB);
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CARD_PULL_CTL1,
+				0x30, 0x10);
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CARD_PULL_CTL5,
+				0x03, 0x01);
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CARD_PULL_CTL6,
+				0x0C, 0x04);
+	}
+
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, SYS_DUMMY0, NYET_MSAK, NYET_EN);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CD_DEGLITCH_WIDTH, 0xFF, 0x08);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD,
+			CD_DEGLITCH_EN, XD_CD_DEGLITCH_EN, 0x0);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, SD30_DRIVE_SEL,
+			SD30_DRIVE_MASK, DRIVER_TYPE_D);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD,
+			CARD_DRIVE_SEL, SD20_DRIVE_MASK, 0x0);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, LDO_POWER_CFG, 0xE0, 0x0);
+
+	if (ucr->is_rts5179)
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD,
+				CARD_PULL_CTL5, 0x03, 0x01);
+
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CARD_DMA1_CTL,
+		       EXTEND_DMA1_ASYNC_SIGNAL, EXTEND_DMA1_ASYNC_SIGNAL);
+	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CARD_INT_PEND,
+			XD_INT | MS_INT | SD_INT,
+			XD_INT | MS_INT | SD_INT);
+
+	ret = rtsx_usb_send_cmd(ucr, MODE_C, 100);
+	if (ret)
+		return ret;
+
+	/* config non-crystal mode */
+	rtsx_usb_read_register(ucr, CFG_MODE, &val);
+	if ((val & XTAL_FREE) || ((val & CLK_MODE_MASK) == CLK_MODE_NON_XTAL)) {
+		ret = rtsx_usb_write_phy_register(ucr, 0xC2, 0x7C);
+		if (ret)
+			return ret;
+	}
+
+	return 0;

drivers/mfd/rtsx_usb.c line 577

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+}
+
+static int rtsx_usb_init_chip(struct rtsx_ucr *ucr)
+{
+	int ret;
+	u8 val;
+
+	rtsx_usb_clear_fsm_err(ucr);
+
+	/* power on SSC*/

White space error.

Quoted 11 lines
drivers/mfd/rtsx_usb.c new file +11 -0
+	ret = rtsx_usb_write_register(ucr,
+			FPDCTL, SSC_POWER_MASK, SSC_POWER_ON);
+	if (ret)
+		return ret;
+
+	usleep_range(100, 1000);
+	ret = rtsx_usb_write_register(ucr, CLK_DIV, CLK_CHANGE, 0x00);
+	if (ret)
+		return ret;
+
+	/* determine IC version */

drivers/mfd/rtsx_usb.c line 598

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+	ret = rtsx_usb_read_register(ucr, HW_VERSION, &val);
+	if (ret)
+		return ret;
+
+	ucr->ic_version = val & HW_VER_MASK;
+
+	/* determine package */
+	ret = rtsx_usb_read_register(ucr, CARD_SHARE_MODE, &val);
+	if (ret)
+		return ret;

'/n' here.

Quoted 5 lines
drivers/mfd/rtsx_usb.c new file +5 -0
+	if (val & CARD_SHARE_LQFP_SEL) {
+		ucr->package = LQFP48;
+		dev_dbg(&ucr->pusb_intf->dev, "Package: LQFP48\n");
+	} else {
+		ucr->package = QFN24;

drivers/mfd/rtsx_usb.c line 613

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+		dev_dbg(&ucr->pusb_intf->dev, "Package: QFN24\n");
+	}
+
+	/* determine IC variations */
+	rtsx_usb_read_register(ucr, CFG_MODE_1, &val);
+	if (val & RTS5179) {
+		ucr->is_rts5179 = 1;
+		dev_dbg(&ucr->pusb_intf->dev, "Device is rts5179\n");
+	} else {
+		ucr->is_rts5179 = 0;

These are better as bools I think.

drivers/mfd/rtsx_usb.c line 620

Quoted 7 lines
drivers/mfd/rtsx_usb.c new file +7 -0
+	}
+
+	ret = rtsx_usb_reset_chip(ucr);
+	if (ret)
+		return ret;
+
+	return 0;

Just
return rtsx_usb_reset_chip(ucr);

Quoted 5 lines
drivers/mfd/rtsx_usb.c new file +5 -0
+}
+
+static int rtsx_usb_probe(struct usb_interface *intf,
+			 const struct usb_device_id *id)
+{

drivers/mfd/rtsx_usb.c line 635

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+	struct usb_device *usb_dev = interface_to_usbdev(intf);
+	struct rtsx_ucr *ucr;
+	int i;
+	int ret;
+
+	dev_dbg(&intf->dev,
+		": Realtek USB Card Reader found at bus %03d address %03d\n",
+		 usb_dev->bus->busnum, usb_dev->devnum);
+
+	ucr = kzalloc(sizeof(struct rtsx_ucr), GFP_KERNEL);

s/struct rtsx_ucr/*ucr/

Any reason for not using managed resources?

drivers/mfd/rtsx_usb.c line 646

Quoted 11 lines
drivers/mfd/rtsx_usb.c new file +11 -0
+	if (!ucr)
+		return -ENOMEM;
+
+	ucr->pusb_dev = usb_dev;
+
+	/* alloc coherent buffer */
+	ucr->iobuf = usb_alloc_coherent(ucr->pusb_dev, IOBUF_SIZE,
+			GFP_KERNEL, &ucr->iobuf_dma);
+	if (!ucr->iobuf) {
+		ret = -ENOMEM;
+		goto out_free_chip;

No need to do this if you use managed resources.

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+	}
+
+	/* set USB interface data */
+	usb_set_intfdata(intf, ucr);
+
+	ucr->vendor_id = id->idVendor;
+	ucr->product_id = id->idProduct;
+	ucr->cmd_buf = ucr->rsp_buf = ucr->iobuf;
+
+	mutex_init(&ucr->dev_mutex);

drivers/mfd/rtsx_usb.c line 666

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+
+	ucr->pusb_intf = intf;
+
+	/* initialize */
+	ret = rtsx_usb_init_chip(ucr);
+	if (ret)
+		goto out_init_fail;
+
+	for (i = 0; i < ARRAY_SIZE(rtsx_usb_cells); i++) {
+		rtsx_usb_cells[i].platform_data = &ucr;

You've already put this in your drvdata (or ntfdata, as it's called here). Why do you also need it in platform_data?

Quoted 8 lines
drivers/mfd/rtsx_usb.c new file +8 -0
+		rtsx_usb_cells[i].pdata_size = sizeof(*ucr);
+	}
+	ret = mfd_add_devices(&intf->dev, usb_dev->devnum, rtsx_usb_cells,
+			ARRAY_SIZE(rtsx_usb_cells), NULL, 0, NULL);
+	if (ret)
+		goto out_init_fail;
+	/* initialize USB SG transfer timer */
+	init_timer(&ucr->sg_timer);

drivers/mfd/rtsx_usb.c line 684

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+	setup_timer(&ucr->sg_timer, rtsx_usb_sg_timed_out, (unsigned long) ucr);
+#ifdef CONFIG_PM
+	intf->needs_remote_wakeup = 1;
+	usb_enable_autosuspend(usb_dev);
+#endif
+
+	return 0;
+
+out_init_fail:
+	usb_set_intfdata(ucr->pusb_intf, NULL);

No need to do this, it's taken care of elsewhere.

Quoted 25 lines
drivers/mfd/rtsx_usb.c new file +25 -0
+	usb_free_coherent(ucr->pusb_dev, IOBUF_SIZE, ucr->iobuf,
+			ucr->iobuf_dma);
+
+out_free_chip:
+	kfree(ucr);
+	dev_err(&intf->dev, "%s failed\n", __func__);
+	return ret;
+}
+
+static void rtsx_usb_disconnect(struct usb_interface *intf)
+{
+	struct rtsx_ucr *ucr = (struct rtsx_ucr *)usb_get_intfdata(intf);
+
+	dev_dbg(&intf->dev, "%s called\n", __func__);
+
+	mfd_remove_devices(&intf->dev);
+
+	usb_set_intfdata(ucr->pusb_intf, NULL);
+	usb_free_coherent(ucr->pusb_dev, IOBUF_SIZE, ucr->iobuf,
+			ucr->iobuf_dma);
+
+	kfree(ucr);
+}
+
+#ifdef CONFIG_PM

drivers/mfd/rtsx_usb.c line 719

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+static int rtsx_usb_suspend(struct usb_interface *intf, pm_message_t message)
+{
+	struct rtsx_ucr *ucr =
+		(struct rtsx_ucr *)usb_get_intfdata(intf);
+
+	dev_dbg(&intf->dev, "%s called with pm message 0x%04u\n",
+			__func__, message.event);
+
+	mutex_lock(&ucr->dev_mutex);
+	rtsx_usb_turn_off_led(ucr);

That's it? That's all you do during suspend? :)

drivers/mfd/rtsx_usb.c line 725

Quoted 6 lines
drivers/mfd/rtsx_usb.c new file +6 -0
+	mutex_unlock(&ucr->dev_mutex);
+	return 0;
+}
+
+static int rtsx_usb_resume(struct usb_interface *intf)
+{

Don't you want to turn the LED back on here?

Or will that happen automatically when you write/read to/from it again?

Quoted 16 lines
drivers/mfd/rtsx_usb.c new file +16 -0
+	return 0;
+}
+
+static int rtsx_usb_reset_resume(struct usb_interface *intf)
+{
+	struct rtsx_ucr *ucr =
+		(struct rtsx_ucr *)usb_get_intfdata(intf);
+
+	rtsx_usb_reset_chip(ucr);
+	return 0;
+}
+
+#else /* CONFIG_PM */
+
+#define rtsx_usb_suspend NULL
+#define rtsx_usb_resume NULL

drivers/mfd/rtsx_usb.c line 751

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+#define rtsx_usb_reset_resume NULL
+
+#endif /* CONFIG_PM */
+
+
+static int rtsx_usb_pre_reset(struct usb_interface *intf)
+{
+	struct rtsx_ucr *ucr = (struct rtsx_ucr *)usb_get_intfdata(intf);
+
+	mutex_lock(&ucr->dev_mutex);

Is this normal?

Quoted 20 lines
drivers/mfd/rtsx_usb.c new file +20 -0
+	return 0;
+}
+
+static int rtsx_usb_post_reset(struct usb_interface *intf)
+{
+	struct rtsx_ucr *ucr = (struct rtsx_ucr *)usb_get_intfdata(intf);
+
+	mutex_unlock(&ucr->dev_mutex);
+	return 0;
+}
+
+static struct usb_device_id rtsx_usb_usb_ids[] = {
+	{ USB_DEVICE(0x0BDA, 0x0129) },
+	{ USB_DEVICE(0x0BDA, 0x0139) },
+	{ USB_DEVICE(0x0BDA, 0x0140) },
+	{ }
+};
+
+static struct usb_driver rtsx_usb_driver = {
+	.name =		DRV_NAME_RTSX_USB,

drivers/mfd/rtsx_usb.c line 781

Quoted 10 lines
drivers/mfd/rtsx_usb.c new file +10 -0
+	.probe =	rtsx_usb_probe,
+	.disconnect =	rtsx_usb_disconnect,
+	.suspend =	rtsx_usb_suspend,
+	.resume =	rtsx_usb_resume,
+	.reset_resume =	rtsx_usb_reset_resume,
+	.pre_reset =	rtsx_usb_pre_reset,
+	.post_reset =	rtsx_usb_post_reset,
+	.id_table =	rtsx_usb_usb_ids,
+	.supports_autosuspend = 1,
+	.soft_unbind =	1,

Better if you line up from the ='s I think:

static struct usb_driver rtsx_usb_driver = {

	.name         =	DRV_NAME_RTSX_USB,
	.probe        =	rtsx_usb_probe,
	.disconnect   =	rtsx_usb_disconnect,
	.suspend      =	rtsx_usb_suspend,

<snip>

drivers/mfd/rtsx_usb.c

Quoted 9 lines
drivers/mfd/rtsx_usb.c new file +9 -0
+#include <linux/usb.h>
+
+/* related module names */
+#define RTSX_USB_SD_CARD		0
+#define RTSX_USB_MS_CARD		1
+
+#define DRV_NAME_RTSX_USB		"rtsx_usb"
+#define DRV_NAME_RTSX_USB_SDMMC		"rtsx_usb_sdmmc"
+#define DRV_NAME_RTSX_USB_MS		"rtsx_usb_ms"

Can you just put the names in the correct places please?

drivers/mfd/rtsx_usb.c

Quoted 4 lines
drivers/mfd/rtsx_usb.c new file +4 -0
+/* endpoint numbers */
+#define EP_BULK_OUT		1
+#define EP_BULK_IN		2
+#define EP_INTR_IN		3

I assume these aren't defined anywhere else?

drivers/mfd/rtsx_usb.c

Quoted 9 lines
drivers/mfd/rtsx_usb.c new file +9 -0
+/* data structures */
+struct rtsx_ucr {
+	u16			vendor_id;
+	u16			product_id;
+
+	int			package;
+#define QFN24			0
+#define LQFP48			1
+#define CHECK_PKG(ucr, pkg)	((ucr)->package == (pkg))

Please remove this from the struct.

drivers/mfd/rtsx_usb.c

Quoted 2 lines
drivers/mfd/rtsx_usb.c new file +2 -0
+	u8			ic_version;
+	u8			is_rts5179;

bool

Quoted 15 lines
drivers/mfd/rtsx_usb.c new file +15 -0
+	unsigned int		cur_clk;
+
+	u8			*cmd_buf;
+	unsigned int		cmd_idx;
+	u8			*rsp_buf;
+
+	struct usb_device	*pusb_dev;
+	struct usb_interface	*pusb_intf;
+	struct usb_sg_request	current_sg;
+	unsigned char		*iobuf;
+	dma_addr_t		iobuf_dma;
+
+	struct timer_list	sg_timer;
+	struct mutex		dev_mutex;
+};

<snip>

drivers/mfd/rtsx_usb.c

Quoted 9 lines
drivers/mfd/rtsx_usb.c new file +9 -0
+static inline void rtsx_usb_init_cmd(struct rtsx_ucr *ucr)
+{
+	ucr->cmd_idx = 0;
+	ucr->cmd_buf[0] = 'R';
+	ucr->cmd_buf[1] = 'T';
+	ucr->cmd_buf[2] = 'C';
+	ucr->cmd_buf[3] = 'R';
+	ucr->cmd_buf[PACKET_TYPE] = BATCH_CMD;
+}

Now you have this same hunk of code 3 times in the same patch.

<snip>

Signature · 4 lines
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog
#3 lore raw

Re: [PATCH 1/3] mfd: Add realtek USB card reader driver

[ Sorry, I am coming down with the flu today so I'm doing dorky things
like reviewing review comments. I'm not sure how coherent I am. ]

Quoted 6 lines · Lee Jones

drivers/mfd/rtsx_usb.c line 44

Quoted 3 lines
drivers/mfd/rtsx_usb.c new file +3 -0
+static void rtsx_usb_sg_timed_out(unsigned long data)
+{
+	struct rtsx_ucr *ucr = (struct rtsx_ucr *)data;

What's going to happen when your device runs 64bit?

I'm not sure I understand what you mean here. On linux sizeof(long) is always the same as sizeof(void *).

Quoted 7 lines

drivers/mfd/rtsx_usb.c line 106

Quoted 5 lines
drivers/mfd/rtsx_usb.c new file +5 -0
+	if (cmd_len > IOBUF_SIZE)
+		return -EINVAL;
+
+	if (cmd_len % 4)
+		cmd_len += (4 - cmd_len % 4);

Please document in a comment.

There is a kernel macro for this:

	cmd_len = ALIGN(cmd_len, 4);

	if (cmd_len > IOBUF_SIZE)
		return -EINVAL;
Quoted 6 lines

Review

Quoted 3 lines
(quoted diff) +2 -0
+
+

Extra '/n'

It weirds me out when you mix up '\n' and /n'.

Quoted 5 lines
Quoted 5 lines
drivers/mfd/rtsx_usb.c new file +5 -0
+int rtsx_usb_ep0_write_register(struct rtsx_ucr *ucr, u16 addr,
+		u8 mask, u8 data)
+{
+	u16 value = 0, index = 0;
+
Quoted 9 lines

drivers/mfd/rtsx_usb.c line 196

Quoted 5 lines
drivers/mfd/rtsx_usb.c new file +5 -0
+	value |= 0x03 << 14;
+	value |= addr & 0x3FFF;
+	value = ((value << 8) & 0xFF00) | ((value >> 8) & 0x00FF);
+	index |= (u16)mask;
+	index |= (u16)data << 8;

Lots of random numbers here, please #define for clarity and ease of reading.

The only really random number is the 0x03, but yeah, it would help if that we a define.

	addr |= 0x03 << 14;

	value = __swab16(addr);
	index = mask | (data << 8);
Quoted 11 lines

drivers/mfd/rtsx_usb.c line 635

Quoted 6 lines
drivers/mfd/rtsx_usb.c new file +6 -0
+
+	dev_dbg(&intf->dev,
+		": Realtek USB Card Reader found at bus %03d address %03d\n",
+		 usb_dev->bus->busnum, usb_dev->devnum);
+
+	ucr = kzalloc(sizeof(struct rtsx_ucr), GFP_KERNEL);

s/struct rtsx_ucr/*ucr/

Any reason for not using managed resources?

Roger, he means the devm_kzalloc().

regards,
dan carpenter

#4 lore raw

Re: [PATCH 1/3] mfd: Add realtek USB card reader driver

Quoted 24 lines · Dan Carpenter

[ Sorry, I am coming down with the flu today so I'm doing dorky things
like reviewing review comments. I'm not sure how coherent I am. ]

Quoted 6 lines · Lee Jones

drivers/mfd/rtsx_usb.c line 44

Quoted 3 lines
drivers/mfd/rtsx_usb.c new file +3 -0
+static void rtsx_usb_sg_timed_out(unsigned long data)
+{
+	struct rtsx_ucr *ucr = (struct rtsx_ucr *)data;

What's going to happen when your device runs 64bit?

I'm not sure I understand what you mean here. On linux sizeof(long) is always the same as sizeof(void *).

Quoted 7 lines

drivers/mfd/rtsx_usb.c line 106

Quoted 5 lines
drivers/mfd/rtsx_usb.c new file +5 -0
+	if (cmd_len > IOBUF_SIZE)
+		return -EINVAL;
+
+	if (cmd_len % 4)
+		cmd_len += (4 - cmd_len % 4);

Please document in a comment.

Quoted 4 lines

There is a kernel macro for this:

	cmd_len = ALIGN(cmd_len, 4);

Btw, what's the difference between ALIGN(cmd_len, 4) and round_up(cmd_len, 4)? Maybe round_up() is better.

regards,
dan carpenter

#5 lore raw

Re: [PATCH 1/3] mfd: Add realtek USB card reader driver

In reply to Dan Carpenter
Quoted 2 lines

[ Sorry, I am coming down with the flu today so I'm doing dorky things
like reviewing review comments. I'm not sure how coherent I am. ]

Always welcome.

NB: I did this review in double-quick time, which may account for some of the weird thought processes (or lack there of).

Quoted 9 lines
Quoted 6 lines

drivers/mfd/rtsx_usb.c line 44

Quoted 3 lines
drivers/mfd/rtsx_usb.c new file +3 -0
+static void rtsx_usb_sg_timed_out(unsigned long data)
+{
+	struct rtsx_ucr *ucr = (struct rtsx_ucr *)data;

What's going to happen when your device runs 64bit?

I'm not sure I understand what you mean here. On linux sizeof(long) is always the same as sizeof(void *).

I had an odd moment where I thought we'd need long long for 64bit.

Nevermind.

Quoted 8 lines
Quoted 7 lines

drivers/mfd/rtsx_usb.c line 106

Quoted 5 lines
drivers/mfd/rtsx_usb.c new file +5 -0
+	if (cmd_len > IOBUF_SIZE)
+		return -EINVAL;
+
+	if (cmd_len % 4)
+		cmd_len += (4 - cmd_len % 4);

Please document in a comment.

Quoted 6 lines

There is a kernel macro for this:

	cmd_len = ALIGN(cmd_len, 4);

	if (cmd_len > IOBUF_SIZE)
		return -EINVAL;

I had a feeling this was the intention, hence why I was asking for a comment. But yes, if all this is doing is alignment then the kernel macro is preferred.

Quoted 7 lines
Quoted 5 lines

Review

Quoted 2 lines
(quoted diff) +2 -0
+
+

Extra '/n'

It weirds me out when you mix up '\n' and /n'.

Typos happen on occasion...

Quoted 15 lines
Quoted 5 lines
Quoted 5 lines
drivers/mfd/rtsx_usb.c new file +5 -0
+int rtsx_usb_ep0_write_register(struct rtsx_ucr *ucr, u16 addr,
+		u8 mask, u8 data)
+{
+	u16 value = 0, index = 0;
+
Quoted 9 lines

drivers/mfd/rtsx_usb.c line 196

Quoted 5 lines
drivers/mfd/rtsx_usb.c new file +5 -0
+	value |= 0x03 << 14;
+	value |= addr & 0x3FFF;
+	value = ((value << 8) & 0xFF00) | ((value >> 8) & 0x00FF);
+	index |= (u16)mask;
+	index |= (u16)data << 8;

Lots of random numbers here, please #define for clarity and ease of reading.

Quoted 2 lines

The only really random number is the 0x03, but yeah, it would help if that we a define.

See ---^

As I say, I just glossed over the code, thus didn't spend the time to work out the arithmetic. Now I do, most of it is pretty self explanatory. I would also like to see the SHIFT #defined, although this may become superfluous once the 0x03 is clarified.

Quoted 4 lines
	addr |= 0x03 << 14;

	value = __swab16(addr);
	index = mask | (data << 8);

This is 100% better/clearer.

Quoted 9 lines
Quoted 9 lines

drivers/mfd/rtsx_usb.c line 635

Quoted 6 lines
drivers/mfd/rtsx_usb.c new file +6 -0
+
+	dev_dbg(&intf->dev,
+		": Realtek USB Card Reader found at bus %03d address %03d\n",
+		 usb_dev->bus->busnum, usb_dev->devnum);
+
+	ucr = kzalloc(sizeof(struct rtsx_ucr), GFP_KERNEL);

s/struct rtsx_ucr/*ucr/

Quoted 3 lines
Quoted 2 lines

Any reason for not using managed resources?

Roger, he means the devm_kzalloc().

That I do.

Signature · 8 lines
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog
_______________________________________________
devel mailing list
[email protected]
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
author
#6 lore raw

Re: [PATCH 1/3] mfd: Add realtek USB card reader driver

In reply to Lee Jones
Quoted 7 lines · Lee Jones
Quoted 7 lines

From: Roger Tseng <[email protected]>

Realtek USB card reader provides a channel to transfer command or data to flash memory cards. This driver exports host instances for mmc and memstick subsystems and handles basic works.

[snip]

Quoted 17 lines
Quoted 17 lines
drivers/mfd/rtsx_usb.c new file +17 -0
+static int rtsx_usb_bulk_transfer_sglist(struct rtsx_ucr *ucr,
+             unsigned int pipe, struct scatterlist *sg, int num_sg,
+             unsigned int length, unsigned int *act_len, int timeout)
+{
+     int ret;
+
+     dev_dbg(&ucr->pusb_intf->dev, "%s: xfer %u bytes, %d entries\n",
+                     __func__, length, num_sg);
+     ret = usb_sg_init(&ucr->current_sg, ucr->pusb_dev, pipe, 0,
+                     sg, num_sg, length, GFP_NOIO);
+     if (ret)
+             return ret;
+
+     ucr->sg_timer.expires = jiffies + msecs_to_jiffies(timeout);
+     add_timer(&ucr->sg_timer);
+     usb_sg_wait(&ucr->current_sg);
+     del_timer(&ucr->sg_timer);
Quoted 9 lines

drivers/mfd/rtsx_usb.c line 75

Quoted 6 lines
drivers/mfd/rtsx_usb.c new file +6 -0
+
+     if (act_len)
+             *act_len = ucr->current_sg.bytes;
+
+     return ucr->current_sg.status;
+}

Code looks fine, but shouldn't this live an a USB driver?

We have studied drivers in usb/storage, the place that most likely to put the driver. All of them are for STANDARD usb mass storage class(something like USB_CLASS_MASS_STORAGE). But this is not the same case. This driver is for our vendor-class device with completely different protocol. It is really an USB interfaced flash card command converter(or channel) device rather than a real storage. It also has two derived modules(rtsx_usb_sdmmc, rtsx_usb_memstick) for other two subsystems.

We also have another driver: rtsx_pcr.c resident in mfd/ for similar devices but of PCIe interface. It is nature for us to choose the same place for this variant.

[snip]

Quoted 15 lines
Quoted 15 lines
drivers/mfd/rtsx_usb.c new file +15 -0
+     }
+
+     /* set USB interface data */
+     usb_set_intfdata(intf, ucr);
+
+     ucr->vendor_id = id->idVendor;
+     ucr->product_id = id->idProduct;
+     ucr->cmd_buf = ucr->rsp_buf = ucr->iobuf;
+
+     mutex_init(&ucr->dev_mutex);
+
+     ucr->pusb_intf = intf;
+
+     /* initialize */
+     ret = rtsx_usb_init_chip(ucr);
Quoted 8 lines

drivers/mfd/rtsx_usb.c line 666

Quoted 5 lines
drivers/mfd/rtsx_usb.c new file +5 -0
+     if (ret)
+             goto out_init_fail;
+
+     for (i = 0; i < ARRAY_SIZE(rtsx_usb_cells); i++) {
+             rtsx_usb_cells[i].platform_data = &ucr;

You've already put this in your drvdata (or ntfdata, as it's called here). Why do you also need it in platform_data?

Derived modules rtsx_usb_sdmmc and rtsx_usb_memstick will retrieve this from platform_data. They will not be aware of usb interface struct.

[snip]

Quoted 5 lines
Quoted 5 lines
drivers/mfd/rtsx_usb.c new file +5 -0
+#ifdef CONFIG_PM
+static int rtsx_usb_suspend(struct usb_interface *intf, pm_message_t message)
+{
+     struct rtsx_ucr *ucr =
+             (struct rtsx_ucr *)usb_get_intfdata(intf);
Quoted 9 lines

drivers/mfd/rtsx_usb.c line 719

Quoted 6 lines
drivers/mfd/rtsx_usb.c new file +6 -0
+
+     dev_dbg(&intf->dev, "%s called with pm message 0x%04u\n",
+                     __func__, message.event);
+
+     mutex_lock(&ucr->dev_mutex);
+     rtsx_usb_turn_off_led(ucr);

That's it? That's all you do during suspend? :)

Yes. The rest of power things like turning-off card power or clock should be taken care by cores of derived modules(mmc, memstick). We put only one task here to make sure the LED is off, preventing any module from turn it on but doesn't turn off before suspend.

Quoted 11 lines

drivers/mfd/rtsx_usb.c line 725

Quoted 6 lines
drivers/mfd/rtsx_usb.c new file +6 -0
+     mutex_unlock(&ucr->dev_mutex);
+     return 0;
+}
+
+static int rtsx_usb_resume(struct usb_interface *intf)
+{

Don't you want to turn the LED back on here?

Or will that happen automatically when you write/read to/from it again?

Yes, it would blink again during use. The turn-off before suspend is not a permanent power off.

Quoted 20 lines
Quoted 20 lines
drivers/mfd/rtsx_usb.c new file +20 -0
+     return 0;
+}
+
+static int rtsx_usb_reset_resume(struct usb_interface *intf)
+{
+     struct rtsx_ucr *ucr =
+             (struct rtsx_ucr *)usb_get_intfdata(intf);
+
+     rtsx_usb_reset_chip(ucr);
+     return 0;
+}
+
+#else /* CONFIG_PM */
+
+#define rtsx_usb_suspend NULL
+#define rtsx_usb_resume NULL
+#define rtsx_usb_reset_resume NULL
+
+#endif /* CONFIG_PM */
+
Quoted 8 lines

drivers/mfd/rtsx_usb.c line 751

Quoted 6 lines
drivers/mfd/rtsx_usb.c new file +6 -0
+
+static int rtsx_usb_pre_reset(struct usb_interface *intf)
+{
+     struct rtsx_ucr *ucr = (struct rtsx_ucr *)usb_get_intfdata(intf);
+
+     mutex_lock(&ucr->dev_mutex);

Is this normal?

Yes. It is used to prevent commands during port reset. Same to the one in usb/storage/usb.c.

[snip]

Quoted 13 lines

include/linux/mfd/rtsx_usb.h line 33

Quoted 10 lines
include/linux/mfd/rtsx_usb.h new file +9 -0
+#include <linux/usb.h>
+
+/* related module names */
+#define RTSX_USB_SD_CARD             0
+#define RTSX_USB_MS_CARD             1
+
+#define DRV_NAME_RTSX_USB            "rtsx_usb"
+#define DRV_NAME_RTSX_USB_SDMMC              "rtsx_usb_sdmmc"
+#define DRV_NAME_RTSX_USB_MS         "rtsx_usb_ms"

Can you just put the names in the correct places please?

Do you mean just remove these definitions and fill the string directly at the using place?

Quoted 7 lines

include/linux/mfd/rtsx_usb.h line 38

Quoted 4 lines
include/linux/mfd/rtsx_usb.h new file +4 -0
+/* endpoint numbers */
+#define EP_BULK_OUT          1
+#define EP_BULK_IN           2
+#define EP_INTR_IN           3

I assume these aren't defined anywhere else?

It should not be defined anywhere else since it is really depends on the hardware design but not any general spec.

[snip]

Quoted 8 lines

--
Lee Jones
Linaro STMicroelectronics Landing Team Lead

Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog

------Please consider the environment before printing this e-mail.

#7 lore raw

Re: [PATCH 1/3] mfd: Add realtek USB card reader driver

Quoted 35 lines
Quoted 17 lines
Quoted 17 lines
drivers/mfd/rtsx_usb.c new file +17 -0
+static int rtsx_usb_bulk_transfer_sglist(struct rtsx_ucr *ucr,
+             unsigned int pipe, struct scatterlist *sg, int num_sg,
+             unsigned int length, unsigned int *act_len, int timeout)
+{
+     int ret;
+
+     dev_dbg(&ucr->pusb_intf->dev, "%s: xfer %u bytes, %d entries\n",
+                     __func__, length, num_sg);
+     ret = usb_sg_init(&ucr->current_sg, ucr->pusb_dev, pipe, 0,
+                     sg, num_sg, length, GFP_NOIO);
+     if (ret)
+             return ret;
+
+     ucr->sg_timer.expires = jiffies + msecs_to_jiffies(timeout);
+     add_timer(&ucr->sg_timer);
+     usb_sg_wait(&ucr->current_sg);
+     del_timer(&ucr->sg_timer);
Quoted 9 lines

drivers/mfd/rtsx_usb.c line 75

Quoted 6 lines
drivers/mfd/rtsx_usb.c new file +6 -0
+
+     if (act_len)
+             *act_len = ucr->current_sg.bytes;
+
+     return ucr->current_sg.status;
+}

Code looks fine, but shouldn't this live in a USB driver?

We have studied drivers in usb/storage, the place that most likely to put the driver. All of them are for STANDARD usb mass storage class(something like USB_CLASS_MASS_STORAGE). But this is not the same case. This driver is for our vendor-class device with completely different protocol. It is really an USB interfaced flash card command converter(or channel) device rather than a real storage. It also has two derived modules(rtsx_usb_sdmmc, rtsx_usb_memstick) for other two subsystems.

Quoted 3 lines

We also have another driver: rtsx_pcr.c resident in mfd/ for similar devices but of PCIe interface. It is nature for us to choose the same place for this variant.

Very well, as long as it truly does not fit in drivers/usb. It would be good to have one of the USB folk give the nod though.

Quoted 24 lines
Quoted 15 lines
Quoted 15 lines
drivers/mfd/rtsx_usb.c new file +15 -0
+     }
+
+     /* set USB interface data */
+     usb_set_intfdata(intf, ucr);
+
+     ucr->vendor_id = id->idVendor;
+     ucr->product_id = id->idProduct;
+     ucr->cmd_buf = ucr->rsp_buf = ucr->iobuf;
+
+     mutex_init(&ucr->dev_mutex);
+
+     ucr->pusb_intf = intf;
+
+     /* initialize */
+     ret = rtsx_usb_init_chip(ucr);
Quoted 8 lines

drivers/mfd/rtsx_usb.c line 666

Quoted 5 lines
drivers/mfd/rtsx_usb.c new file +5 -0
+     if (ret)
+             goto out_init_fail;
+
+     for (i = 0; i < ARRAY_SIZE(rtsx_usb_cells); i++) {
+             rtsx_usb_cells[i].platform_data = &ucr;

You've already put this in your drvdata (or ntfdata, as it's called here). Why do you also need it in platform_data?

Quoted 3 lines

Derived modules rtsx_usb_sdmmc and rtsx_usb_memstick will retrieve this from platform_data. They will not be aware of usb interface struct.

They don't need to. If the MMC or Memstick subsystems do not have their own helpers you can just use something like:

  struct rtsx_ucr *ucr = dev_get_drvdata(pdev->dev.parent);

Naturally this is untested code and might not work off the bat, but it does give you some idea of what you can do without iterating through and populating each sub-device's platform data.

Quoted 8 lines
Quoted 6 lines

include/linux/mfd/rtsx_usb.h line 33

Quoted 3 lines
include/linux/mfd/rtsx_usb.h new file +3 -0
+#define DRV_NAME_RTSX_USB            "rtsx_usb"
+#define DRV_NAME_RTSX_USB_SDMMC              "rtsx_usb_sdmmc"
+#define DRV_NAME_RTSX_USB_MS         "rtsx_usb_ms"

Can you just put the names in the correct places please?

Do you mean just remove these definitions and fill the string directly at the using place?

I do.

I find these types of defines unhelpful and somewhat obfuscating.

Signature · 8 lines
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog
_______________________________________________
devel mailing list
[email protected]
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
author
#8 lore raw

Re: [PATCH 1/3] mfd: Add realtek USB card reader driver

Quoted 35 lines · Lee Jones
Quoted 35 lines
Quoted 17 lines
Quoted 17 lines
drivers/mfd/rtsx_usb.c new file +17 -0
+static int rtsx_usb_bulk_transfer_sglist(struct rtsx_ucr *ucr,
+             unsigned int pipe, struct scatterlist *sg, int num_sg,
+             unsigned int length, unsigned int *act_len, int timeout)
+{
+     int ret;
+
+     dev_dbg(&ucr->pusb_intf->dev, "%s: xfer %u bytes, %d entries\n",
+                     __func__, length, num_sg);
+     ret = usb_sg_init(&ucr->current_sg, ucr->pusb_dev, pipe, 0,
+                     sg, num_sg, length, GFP_NOIO);
+     if (ret)
+             return ret;
+
+     ucr->sg_timer.expires = jiffies + msecs_to_jiffies(timeout);
+     add_timer(&ucr->sg_timer);
+     usb_sg_wait(&ucr->current_sg);
+     del_timer(&ucr->sg_timer);
Quoted 9 lines

drivers/mfd/rtsx_usb.c line 75

Quoted 6 lines
drivers/mfd/rtsx_usb.c new file +6 -0
+
+     if (act_len)
+             *act_len = ucr->current_sg.bytes;
+
+     return ucr->current_sg.status;
+}

Code looks fine, but shouldn't this live in a USB driver?

We have studied drivers in usb/storage, the place that most likely to put the driver. All of them are for STANDARD usb mass storage class(something like USB_CLASS_MASS_STORAGE). But this is not the same case. This driver is for our vendor-class device with completely different protocol. It is really an USB interfaced flash card command converter(or channel) device rather than a real storage. It also has two derived modules(rtsx_usb_sdmmc, rtsx_usb_memstick) for other two subsystems.

Quoted 7 lines
Quoted 3 lines

We also have another driver: rtsx_pcr.c resident in mfd/ for similar devices but of PCIe interface. It is nature for us to choose the same place for this variant.

Very well, as long as it truly does not fit in drivers/usb. It would be good to have one of the USB folk give the nod though.

I found Greg K-H is exactly one of the maintainers of USB subsystem as I read the MAINTAINERS document.

Greg, would you please comment this subsystem concern or introduce other members? Thanks.

#9 lore raw

Re: [PATCH 1/3] mfd: Add realtek USB card reader driver

Quoted 40 lines · Roger
Quoted 39 lines · Lee Jones
Quoted 38 lines
Quoted 17 lines
Quoted 17 lines
drivers/mfd/rtsx_usb.c new file +17 -0
+static int rtsx_usb_bulk_transfer_sglist(struct rtsx_ucr *ucr,
+             unsigned int pipe, struct scatterlist *sg, int num_sg,
+             unsigned int length, unsigned int *act_len, int timeout)
+{
+     int ret;
+
+     dev_dbg(&ucr->pusb_intf->dev, "%s: xfer %u bytes, %d entries\n",
+                     __func__, length, num_sg);
+     ret = usb_sg_init(&ucr->current_sg, ucr->pusb_dev, pipe, 0,
+                     sg, num_sg, length, GFP_NOIO);
+     if (ret)
+             return ret;
+
+     ucr->sg_timer.expires = jiffies + msecs_to_jiffies(timeout);
+     add_timer(&ucr->sg_timer);
+     usb_sg_wait(&ucr->current_sg);
+     del_timer(&ucr->sg_timer);
Quoted 9 lines

drivers/mfd/rtsx_usb.c line 75

Quoted 6 lines
drivers/mfd/rtsx_usb.c new file +6 -0
+
+     if (act_len)
+             *act_len = ucr->current_sg.bytes;
+
+     return ucr->current_sg.status;
+}

Code looks fine, but shouldn't this live in a USB driver?

We have studied drivers in usb/storage, the place that most likely to put the driver. All of them are for STANDARD usb mass storage class(something like USB_CLASS_MASS_STORAGE). But this is not the same case. This driver is for our vendor-class device with completely different protocol. It is really an USB interfaced flash card command converter(or channel) device rather than a real storage. It also has two derived modules(rtsx_usb_sdmmc, rtsx_usb_memstick) for other two subsystems.

We also have another driver: rtsx_pcr.c resident in mfd/ for similar devices but of PCIe interface. It is nature for us to choose the same place for this variant.

Quoted 8 lines
Quoted 3 lines

Very well, as long as it truly does not fit in drivers/usb. It would be good to have one of the USB folk give the nod though.

I found Greg K-H is exactly one of the maintainers of USB subsystem as I read the MAINTAINERS document.

Greg, would you please comment this subsystem concern or introduce other members? Thanks.

It's the middle of the merge window, nothing can happen until after 3.14-rc1 is out. So how about resending all of this in two weeks after that happens so we can all properly discuss it?

As for where the driver should be located, if it shares logic with the usb storage driver, then it should be in drivers/usb/storage, otherwise put it wherever makes most sense, there's no need to put all USB drivers under drivers/usb/ at all.

thanks,

greg k-h

#10 lore raw

Re: [PATCH 1/3] mfd: Add realtek USB card reader driver

In reply to [email protected]
Quoted 55 lines · On Tue, Jan 14, 2014 at 03:47:34PM +0800, [email protected] wrote:

From: Roger Tseng <[email protected]>

Realtek USB card reader provides a channel to transfer command or data to flash memory cards. This driver exports host instances for mmc and memstick subsystems and handles basic works.

---
 drivers/mfd/Kconfig          |  10 +
 drivers/mfd/Makefile         |   1 +
 drivers/mfd/rtsx_usb.c       | 788 +++++++++++++++++++++++++++++++++++++++++++
 include/linux/mfd/rtsx_usb.h | 619 +++++++++++++++++++++++++++++++++
 4 files changed, 1418 insertions(+)
 create mode 100644 drivers/mfd/rtsx_usb.c
 create mode 100644 include/linux/mfd/rtsx_usb.h
drivers/mfd/Kconfig +10 -0
index b7c74a7..4c99ebd 100644
@@ -507,6 +507,16 @@ config MFD_RTSX_PCI
 	  types of memory cards, such as Memory Stick, Memory Stick Pro,
 	  Secure Digital and MultiMediaCard.
 
+config MFD_RTSX_USB
+	tristate "Realtek USB card reader"
+	depends on USB
+	select MFD_CORE
+	help
+	Select this option to get support for Realtek USB 2.0 card readers
+	including RTS5129, RTS5139, RTS5179 and RTS5170.
+	Realtek card reader supports access to many types of memory cards,
+	such as Memory Stick Pro, Secure Digital and MultiMediaCard.
+
 config MFD_RC5T583
 	bool "Ricoh RC5T583 Power Management system device"
 	depends on I2C=y
drivers/mfd/Makefile +1 -0
index 8a28dc9..33b8de6 100644
@@ -14,6 +14,7 @@ obj-$(CONFIG_MFD_CROS_EC_SPI)	+= cros_ec_spi.o
 
 rtsx_pci-objs			:= rtsx_pcr.o rts5209.o rts5229.o rtl8411.o rts5227.o rts5249.o
 obj-$(CONFIG_MFD_RTSX_PCI)	+= rtsx_pci.o
+obj-$(CONFIG_MFD_RTSX_USB)	+= rtsx_usb.o
 
 obj-$(CONFIG_HTC_EGPIO)		+= htc-egpio.o
 obj-$(CONFIG_HTC_PASIC3)	+= htc-pasic3.o
drivers/mfd/rtsx_usb.c new file +0 -0
new file mode 100644
index 0000000..905ec8d

drivers/mfd/rtsx_usb.c line 8

Quoted 8 lines
drivers/mfd/rtsx_usb.c new file +8 -0
@@ -0,0 +1,788 @@
+/* Driver for Realtek USB card reader
+ *
+ * Copyright(c) 2009-2013 Realtek Semiconductor Corp. All rights reserved.
+ *
+ * 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, or (at your option) any
+ * later version.

Do you really mean "any later version"? (sorry, I have to ask.)

Same goes for the other files you add with this specific license wording.

drivers/mfd/rtsx_usb.c line 27

Quoted 7 lines
drivers/mfd/rtsx_usb.c new file +7 -0
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/mutex.h>
+#include <linux/usb.h>
+#include <linux/platform_device.h>
+#include <linux/mfd/core.h>
+#include <asm/unaligned.h>

Why is this file neded?

thanks,

greg k-h

#11 lore raw

PATCH v2 2/3 mmc: Add realtek USB sdmmc host driver

In reply to [email protected]

From: Roger Tseng <[email protected]>

Realtek USB SD/MMC host driver provides mmc host support based on the Realtek USB card reader MFD driver.

---
 drivers/mmc/host/Kconfig          |    7 +
 drivers/mmc/host/Makefile         |    1 +
 drivers/mmc/host/rtsx_usb_sdmmc.c | 1509 +++++++++++++++++++++++++++++++++++++
 3 files changed, 1517 insertions(+)
 create mode 100644 drivers/mmc/host/rtsx_usb_sdmmc.c
drivers/mmc/host/Kconfig +7 -0
index 7fc5099..16f9a23 100644
@@ -665,3 +665,10 @@ config MMC_REALTEK_PCI
 	help
 	  Say Y here to include driver code to support SD/MMC card interface
 	  of Realtek PCI-E card reader
+
+config MMC_REALTEK_USB
+	tristate "Realtek USB SD/MMC Card Interface Driver"
+	depends on MFD_RTSX_USB
+	help
+	  Say Y here to include driver code to support SD/MMC card interface
+	  of Realtek RTS5129/39 series card reader
drivers/mmc/host/Makefile +1 -0
index c41d0c3..9d8d765 100644
@@ -51,6 +51,7 @@ obj-$(CONFIG_MMC_USHC)		+= ushc.o
 obj-$(CONFIG_MMC_WMT)		+= wmt-sdmmc.o
 
 obj-$(CONFIG_MMC_REALTEK_PCI)	+= rtsx_pci_sdmmc.o
+obj-$(CONFIG_MMC_REALTEK_USB)	+= rtsx_usb_sdmmc.o
 
 obj-$(CONFIG_MMC_SDHCI_PLTFM)		+= sdhci-pltfm.o
 obj-$(CONFIG_MMC_SDHCI_CNS3XXX)		+= sdhci-cns3xxx.o
drivers/mmc/host/rtsx_usb_sdmmc.c new file +1509 -0

1,509 lines not on this page · raw message

Signature · 1 line
1.8.2
#12 lore raw

PATCH v2 3/3 memstick: Add realtek USB memstick host driver

In reply to [email protected]

From: Roger Tseng <[email protected]>

Realtek USB memstick host driver provides memstick host support based on the Realtek USB card reader MFD driver.

---
 drivers/memstick/host/Kconfig       |  10 +
 drivers/memstick/host/Makefile      |   1 +
 drivers/memstick/host/rtsx_usb_ms.c | 855 ++++++++++++++++++++++++++++++++++++
 3 files changed, 866 insertions(+)
 create mode 100644 drivers/memstick/host/rtsx_usb_ms.c
drivers/memstick/host/Kconfig +10 -0
index 1b37cf8..7310e32 100644
@@ -52,3 +52,13 @@ config MEMSTICK_REALTEK_PCI
 
 	  To compile this driver as a module, choose M here: the module will
 	  be called rtsx_pci_ms.
+
+config MEMSTICK_REALTEK_USB
+	tristate "Realtek USB Memstick Card Interface Driver"
+	depends on MFD_RTSX_USB
+	help
+	  Say Y here to include driver code to support Memstick card interface
+	  of Realtek RTS5129/39 series USB card reader
+
+	  To compile this driver as a module, choose M here: the module will
+	  be called rts5139_ms.
drivers/memstick/host/Makefile +1 -0
index af3459d..491c955 100644
@@ -6,3 +6,4 @@ obj-$(CONFIG_MEMSTICK_TIFM_MS)		+= tifm_ms.o
 obj-$(CONFIG_MEMSTICK_JMICRON_38X)	+= jmb38x_ms.o
 obj-$(CONFIG_MEMSTICK_R592)		+= r592.o
 obj-$(CONFIG_MEMSTICK_REALTEK_PCI)	+= rtsx_pci_ms.o
+obj-$(CONFIG_MEMSTICK_REALTEK_USB)	+= rtsx_usb_ms.o
drivers/memstick/host/rtsx_usb_ms.c new file +855 -0

855 lines not on this page · raw message

Signature · 1 line
1.8.2
#13 lore raw

Re: [PATCH 3/3] memstick: Add realtek USB memstick host driver

Quoted 22 lines · On Tue, Jan 14, 2014 at 03:47:36PM +0800, [email protected] wrote:
drivers/memstick/host/rtsx_usb_ms.c new file +22 -0
+static int ms_pull_ctl_disable(struct rtsx_ucr *ucr)
+{
+	rtsx_usb_init_cmd(ucr);
+
+	if (CHECK_PKG(ucr, LQFP48)) {
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD,
+				CARD_PULL_CTL1, 0xFF, 0x55);
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD,
+				CARD_PULL_CTL2, 0xFF, 0x55);
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD,
+				CARD_PULL_CTL3, 0xFF, 0x95);
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD,
+				CARD_PULL_CTL4, 0xFF, 0x55);
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD,
+				CARD_PULL_CTL5, 0xFF, 0x55);
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD,
+				CARD_PULL_CTL6, 0xFF, 0xA5);
+	} else {
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD,
+				CARD_PULL_CTL1, 0xFF, 0x65);
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD,
+				CARD_PULL_CTL2, 0xFF, 0x55);

drivers/memstick/host/rtsx_usb_ms.c line 150

Quoted 10 lines
drivers/memstick/host/rtsx_usb_ms.c new file +10 -0
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD,
+				CARD_PULL_CTL3, 0xFF, 0x95);
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD,
+				CARD_PULL_CTL4, 0xFF, 0x55);
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD,
+				CARD_PULL_CTL5, 0xFF, 0x56);
+		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD,
+				CARD_PULL_CTL6, 0xFF, 0x59);
+	}
+

I'm fine with this patch going in as-is, but I just wanted to comment on this for future reference/cleanups. These blocks where all the lines are over 80 characters are hard to look at. We could break it into separate functions.

static int ms_pull_ctl_disable_lqfp48(struct rtsx_ucr *ucr)

{
	rtsx_usb_init_cmd(ucr);

	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CARD_PULL_CTL1, 0xFF, 0x55);
	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CARD_PULL_CTL2, 0xFF, 0x55);
	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CARD_PULL_CTL3, 0xFF, 0x95);
	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CARD_PULL_CTL4, 0xFF, 0x55);
	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CARD_PULL_CTL5, 0xFF, 0x55);
	rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, CARD_PULL_CTL6, 0xFF, 0xA5);

	return rtsx_usb_send_cmd(ucr, MODE_C, 100);
}

Or another option would be to remove the "CARD_" prefix. I don't think we would miss it.

	if (CHECK_PKG(ucr, LQFP48)) {
		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, PULL_CTL1, 0xFF, 0x55);
		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, PULL_CTL2, 0xFF, 0x55);
		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, PULL_CTL3, 0xFF, 0x95);
		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, PULL_CTL4, 0xFF, 0x55);
		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, PULL_CTL5, 0xFF, 0x55);
		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, PULL_CTL6, 0xFF, 0xA5);
	} else {
		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, PULL_CTL1, 0xFF, 0x65);
		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, PULL_CTL2, 0xFF, 0x55);
		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, PULL_CTL3, 0xFF, 0x95);
		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, PULL_CTL4, 0xFF, 0x55);
		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, PULL_CTL5, 0xFF, 0x56);
		rtsx_usb_add_cmd(ucr, WRITE_REG_CMD, PULL_CTL6, 0xFF, 0x59);
	}

Another option might be to remove the "_usb" from the rtsx_usb_add_cmd().

Just some ideas.

regards,
dan carpenter