[U-Boot] [PATCH 1/8] ums: support block devices not MMC devices

[U-Boot] [PATCH 1/8] ums: support block devices not MMC devices

[Stephen Warren]

From: Stephen Warren <swarren@nvidia.com>

The USB Mass Storage function could equally well support a SATA device
as support an MMC device. Update struct ums to contain a block device
descriptor, not an MMC device descriptor.

Cc: Przemyslaw Marczak <p.marczak@samsung.com>
Cc: Lukasz Majewski <l.majewski@samsung.com>
Signed-off-by: Stephen Warren <swarren@nvidia.com>
---
 board/samsung/common/ums.c | 7 ++++---
 include/usb_mass_storage.h | 4 ++--
 2 files changed, 6 insertions(+), 5 deletions(-)

diff --git a/board/samsung/common/ums.c b/board/samsung/common/ums.c
index dc155ad0e5b3..60e7d2a2bd6e 100644
--- a/board/samsung/common/ums.c
+++ b/board/samsung/common/ums.c
@@ -7,12 +7,13 @@
 
 #include <common.h>
 #include <usb_mass_storage.h>
+#include <mmc.h>
 #include <part.h>
 
 static int ums_read_sector(struct ums *ums_dev,
 			   ulong start, lbaint_t blkcnt, void *buf)
 {
-	block_dev_desc_t *block_dev = &ums_dev->mmc->block_dev;
+	block_dev_desc_t *block_dev = ums_dev->block_dev;
 	lbaint_t blkstart = start + ums_dev->start_sector;
 	int dev_num = block_dev->dev;
 
@@ -22,7 +23,7 @@ static int ums_read_sector(struct ums *ums_dev,
 static int ums_write_sector(struct ums *ums_dev,
 			    ulong start, lbaint_t blkcnt, const void *buf)
 {
-	block_dev_desc_t *block_dev = &ums_dev->mmc->block_dev;
+	block_dev_desc_t *block_dev = ums_dev->block_dev;
 	lbaint_t blkstart = start + ums_dev->start_sector;
 	int dev_num = block_dev->dev;
 
@@ -45,7 +46,7 @@ static struct ums *ums_disk_init(struct mmc *mmc)
 		return NULL;
 	}
 
-	ums_dev.mmc = mmc;
+	ums_dev.block_dev = &mmc->block_dev;
 
 	if (ums_end_sector <= mmc_end_sector) {
 		ums_dev.start_sector = UMS_START_SECTOR;
diff --git a/include/usb_mass_storage.h b/include/usb_mass_storage.h
index 058dcf117404..12fbae447c09 100644
--- a/include/usb_mass_storage.h
+++ b/include/usb_mass_storage.h
@@ -9,7 +9,7 @@
 #define __USB_MASS_STORAGE_H__
 
 #define SECTOR_SIZE		0x200
-#include <mmc.h>
+#include <part.h>
 #include <linux/usb/composite.h>
 
 #ifndef UMS_START_SECTOR
@@ -31,7 +31,7 @@ struct ums {
 	unsigned int start_sector;
 	unsigned int num_sectors;
 	const char *name;
-	struct mmc *mmc;
+	block_dev_desc_t *block_dev;
 };
 
 extern struct ums *ums;
-- 
1.8.1.5

[U-Boot] [PATCH 2/8] ums: remove UMS_{NUM, START}_SECTORS + UMS_START_SECTOR

[Stephen Warren]

From: Stephen Warren <swarren@nvidia.com>

These values aren't set anywhere at present, and hence have no effect.
The concept of a single global offset/number of sectors to expose through
USB Mass Storage doesn't even make sense in the face of multiple storage
devices. Remove these defines to simplify the code.

Cc: Przemyslaw Marczak <p.marczak@samsung.com>
Cc: Lukasz Majewski <l.majewski@samsung.com>
Signed-off-by: Stephen Warren <swarren@nvidia.com>
---
 board/samsung/common/ums.c | 14 ++------------
 include/usb_mass_storage.h |  8 --------
 2 files changed, 2 insertions(+), 20 deletions(-)

diff --git a/board/samsung/common/ums.c b/board/samsung/common/ums.c
index 60e7d2a2bd6e..d78c8e763a5d 100644
--- a/board/samsung/common/ums.c
+++ b/board/samsung/common/ums.c
@@ -39,7 +39,6 @@ static struct ums ums_dev = {
 static struct ums *ums_disk_init(struct mmc *mmc)
 {
 	uint64_t mmc_end_sector = mmc->capacity / SECTOR_SIZE;
-	uint64_t ums_end_sector = UMS_NUM_SECTORS + UMS_START_SECTOR;
 
 	if (!mmc_end_sector) {
 		error("MMC capacity is not valid");
@@ -47,17 +46,8 @@ static struct ums *ums_disk_init(struct mmc *mmc)
 	}
 
 	ums_dev.block_dev = &mmc->block_dev;
-
-	if (ums_end_sector <= mmc_end_sector) {
-		ums_dev.start_sector = UMS_START_SECTOR;
-		if (UMS_NUM_SECTORS)
-			ums_dev.num_sectors = UMS_NUM_SECTORS;
-		else
-			ums_dev.num_sectors = mmc_end_sector - UMS_START_SECTOR;
-	} else {
-		ums_dev.num_sectors = mmc_end_sector;
-		puts("UMS: defined bad disk parameters. Using default.\n");
-	}
+	ums_dev.start_sector = 0;
+	ums_dev.num_sectors = mmc_end_sector;
 
 	printf("UMS: disk start sector: %#x, count: %#x\n",
 	       ums_dev.start_sector, ums_dev.num_sectors);
diff --git a/include/usb_mass_storage.h b/include/usb_mass_storage.h
index 12fbae447c09..6fcc51d1264d 100644
--- a/include/usb_mass_storage.h
+++ b/include/usb_mass_storage.h
@@ -12,14 +12,6 @@
 #include <part.h>
 #include <linux/usb/composite.h>
 
-#ifndef UMS_START_SECTOR
-#define UMS_START_SECTOR	0
-#endif
-
-#ifndef UMS_NUM_SECTORS
-#define UMS_NUM_SECTORS		0
-#endif
-
 /* Wait at maximum 60 seconds for cable connection */
 #define UMS_CABLE_READY_TIMEOUT	60
 
-- 
1.8.1.5

[U-Boot] [PATCH 3/8] ums: remove error-checking of MMC device size

[Stephen Warren]

From: Stephen Warren <swarren@nvidia.com>

There's no reason to believe that an MMC device will incorrectly report
its capacity. Remove error checking of this value from ums_disk_init()
to simplify it.

Cc: Przemyslaw Marczak <p.marczak@samsung.com>
Cc: Lukasz Majewski <l.majewski@samsung.com>
Signed-off-by: Stephen Warren <swarren@nvidia.com>
---
 board/samsung/common/ums.c | 9 +--------
 1 file changed, 1 insertion(+), 8 deletions(-)

diff --git a/board/samsung/common/ums.c b/board/samsung/common/ums.c
index d78c8e763a5d..14e3640e355f 100644
--- a/board/samsung/common/ums.c
+++ b/board/samsung/common/ums.c
@@ -38,16 +38,9 @@ static struct ums ums_dev = {
 
 static struct ums *ums_disk_init(struct mmc *mmc)
 {
-	uint64_t mmc_end_sector = mmc->capacity / SECTOR_SIZE;
-
-	if (!mmc_end_sector) {
-		error("MMC capacity is not valid");
-		return NULL;
-	}
-
 	ums_dev.block_dev = &mmc->block_dev;
 	ums_dev.start_sector = 0;
-	ums_dev.num_sectors = mmc_end_sector;
+	ums_dev.num_sectors = mmc->capacity / SECTOR_SIZE;
 
 	printf("UMS: disk start sector: %#x, count: %#x\n",
 	       ums_dev.start_sector, ums_dev.num_sectors);
-- 
1.8.1.5

[U-Boot] [PATCH 3/8] ums: remove error-checking of MMC device size

[Stephen Warren]

On 04/30/2014 03:13 PM, Stephen Warren wrote:
> From: Stephen Warren <swarren@nvidia.com>
> 
> There's no reason to believe that an MMC device will incorrectly report
> its capacity. Remove error checking of this value from ums_disk_init()
> to simplify it.

Hmm. I guess this check might trigger if you tried to export an SD card
slot over UMS, yet there was no SD card present. Still, patch 6/8 in
this series resolves this, since at least for MMC devices, get_device()
calls mmc_init() internally, and hence returns an error in this case.
This is more direct that checking the device size.

[U-Boot] [PATCH 4/8] ums: remove ums_disk_init()

[Stephen Warren]

From: Stephen Warren <swarren@nvidia.com>

Now that ums_disk_init() is so simple, there's no need for it to be a
separate function. Instead, just add it to the tail end of ums_init().

Cc: Przemyslaw Marczak <p.marczak@samsung.com>
Cc: Lukasz Majewski <l.majewski@samsung.com>
Signed-off-by: Stephen Warren <swarren@nvidia.com>
---
 board/samsung/common/ums.c | 19 +++++++------------
 1 file changed, 7 insertions(+), 12 deletions(-)

diff --git a/board/samsung/common/ums.c b/board/samsung/common/ums.c
index 14e3640e355f..68b9f26cac4e 100644
--- a/board/samsung/common/ums.c
+++ b/board/samsung/common/ums.c
@@ -36,8 +36,14 @@ static struct ums ums_dev = {
 	.name = "UMS disk",
 };
 
-static struct ums *ums_disk_init(struct mmc *mmc)
+struct ums *ums_init(unsigned int dev_num)
 {
+	struct mmc *mmc = NULL;
+
+	mmc = find_mmc_device(dev_num);
+	if (!mmc)
+		return NULL;
+
 	ums_dev.block_dev = &mmc->block_dev;
 	ums_dev.start_sector = 0;
 	ums_dev.num_sectors = mmc->capacity / SECTOR_SIZE;
@@ -47,14 +53,3 @@ static struct ums *ums_disk_init(struct mmc *mmc)
 
 	return &ums_dev;
 }
-
-struct ums *ums_init(unsigned int dev_num)
-{
-	struct mmc *mmc = NULL;
-
-	mmc = find_mmc_device(dev_num);
-	if (!mmc)
-		return NULL;
-
-	return ums_disk_init(mmc);
-}
-- 
1.8.1.5

[U-Boot] [PATCH 5/8] ums: move IO support code to common location

[Stephen Warren]

From: Stephen Warren <swarren@nvidia.com>

There's nothing Samsung-/board-specfic about the implementation of
ums_init(). Move the code into cmd_usb_mass_storage.c, so that it can
be shared by any user of that command.

Cc: Przemyslaw Marczak <p.marczak@samsung.com>
Cc: Lukasz Majewski <l.majewski@samsung.com>
Signed-off-by: Stephen Warren <swarren@nvidia.com>
---
 board/samsung/common/Makefile |  1 -
 board/samsung/common/ums.c    | 55 -------------------------------------------
 common/cmd_usb_mass_storage.c | 46 ++++++++++++++++++++++++++++++++++++
 include/usb_mass_storage.h    |  1 -
 4 files changed, 46 insertions(+), 57 deletions(-)
 delete mode 100644 board/samsung/common/ums.c

diff --git a/board/samsung/common/Makefile b/board/samsung/common/Makefile
index 7d2bb8c4a2f3..41d0cc381494 100644
--- a/board/samsung/common/Makefile
+++ b/board/samsung/common/Makefile
@@ -7,7 +7,6 @@
 
 obj-$(CONFIG_SOFT_I2C_MULTI_BUS) += multi_i2c.o
 obj-$(CONFIG_THOR_FUNCTION) += thor.o
-obj-$(CONFIG_CMD_USB_MASS_STORAGE) += ums.o
 obj-$(CONFIG_MISC_COMMON) += misc.o
 
 ifndef CONFIG_SPL_BUILD
diff --git a/board/samsung/common/ums.c b/board/samsung/common/ums.c
deleted file mode 100644
index 68b9f26cac4e..000000000000
--- a/board/samsung/common/ums.c
+++ /dev/null
@@ -1,55 +0,0 @@
-/*
- * Copyright (C) 2013 Samsung Electronics
- * Lukasz Majewski <l.majewski@samsung.com>
- *
- * SPDX-License-Identifier:	GPL-2.0+
- */
-
-#include <common.h>
-#include <usb_mass_storage.h>
-#include <mmc.h>
-#include <part.h>
-
-static int ums_read_sector(struct ums *ums_dev,
-			   ulong start, lbaint_t blkcnt, void *buf)
-{
-	block_dev_desc_t *block_dev = ums_dev->block_dev;
-	lbaint_t blkstart = start + ums_dev->start_sector;
-	int dev_num = block_dev->dev;
-
-	return block_dev->block_read(dev_num, blkstart, blkcnt, buf);
-}
-
-static int ums_write_sector(struct ums *ums_dev,
-			    ulong start, lbaint_t blkcnt, const void *buf)
-{
-	block_dev_desc_t *block_dev = ums_dev->block_dev;
-	lbaint_t blkstart = start + ums_dev->start_sector;
-	int dev_num = block_dev->dev;
-
-	return block_dev->block_write(dev_num, blkstart, blkcnt, buf);
-}
-
-static struct ums ums_dev = {
-	.read_sector = ums_read_sector,
-	.write_sector = ums_write_sector,
-	.name = "UMS disk",
-};
-
-struct ums *ums_init(unsigned int dev_num)
-{
-	struct mmc *mmc = NULL;
-
-	mmc = find_mmc_device(dev_num);
-	if (!mmc)
-		return NULL;
-
-	ums_dev.block_dev = &mmc->block_dev;
-	ums_dev.start_sector = 0;
-	ums_dev.num_sectors = mmc->capacity / SECTOR_SIZE;
-
-	printf("UMS: disk start sector: %#x, count: %#x\n",
-	       ums_dev.start_sector, ums_dev.num_sectors);
-
-	return &ums_dev;
-}
diff --git a/common/cmd_usb_mass_storage.c b/common/cmd_usb_mass_storage.c
index 2f69a53acc67..5ec5da6d4f04 100644
--- a/common/cmd_usb_mass_storage.c
+++ b/common/cmd_usb_mass_storage.c
@@ -9,9 +9,55 @@
 #include <common.h>
 #include <command.h>
 #include <g_dnl.h>
+#include <mmc.h>
+#include <part.h>
 #include <usb.h>
 #include <usb_mass_storage.h>
 
+static int ums_read_sector(struct ums *ums_dev,
+			   ulong start, lbaint_t blkcnt, void *buf)
+{
+	block_dev_desc_t *block_dev = ums_dev->block_dev;
+	lbaint_t blkstart = start + ums_dev->start_sector;
+	int dev_num = block_dev->dev;
+
+	return block_dev->block_read(dev_num, blkstart, blkcnt, buf);
+}
+
+static int ums_write_sector(struct ums *ums_dev,
+			    ulong start, lbaint_t blkcnt, const void *buf)
+{
+	block_dev_desc_t *block_dev = ums_dev->block_dev;
+	lbaint_t blkstart = start + ums_dev->start_sector;
+	int dev_num = block_dev->dev;
+
+	return block_dev->block_write(dev_num, blkstart, blkcnt, buf);
+}
+
+static struct ums ums_dev = {
+	.read_sector = ums_read_sector,
+	.write_sector = ums_write_sector,
+	.name = "UMS disk",
+};
+
+struct ums *ums_init(unsigned int dev_num)
+{
+	struct mmc *mmc = NULL;
+
+	mmc = find_mmc_device(dev_num);
+	if (!mmc)
+		return NULL;
+
+	ums_dev.block_dev = &mmc->block_dev;
+	ums_dev.start_sector = 0;
+	ums_dev.num_sectors = mmc->capacity / SECTOR_SIZE;
+
+	printf("UMS: disk start sector: %#x, count: %#x\n",
+	       ums_dev.start_sector, ums_dev.num_sectors);
+
+	return &ums_dev;
+}
+
 int do_usb_mass_storage(cmd_tbl_t *cmdtp, int flag,
 			       int argc, char * const argv[])
 {
diff --git a/include/usb_mass_storage.h b/include/usb_mass_storage.h
index 6fcc51d1264d..ce0af27d51cf 100644
--- a/include/usb_mass_storage.h
+++ b/include/usb_mass_storage.h
@@ -30,7 +30,6 @@ extern struct ums *ums;
 
 int fsg_init(struct ums *);
 void fsg_cleanup(void);
-struct ums *ums_init(unsigned int);
 int fsg_main_thread(void *);
 
 #ifdef CONFIG_USB_GADGET_MASS_STORAGE
-- 
1.8.1.5

[U-Boot] [PATCH 6/8] ums: use get_device() not find_mmc_device();

[Stephen Warren]

From: Stephen Warren <swarren@nvidia.com>

get_device() is a generic routine that will support any type of block
device. Use this instead of the type-specific find_mmc_device(), for
future flexibility.

Cc: Przemyslaw Marczak <p.marczak@samsung.com>
Cc: Lukasz Majewski <l.majewski@samsung.com>
Signed-off-by: Stephen Warren <swarren@nvidia.com>
---
 common/cmd_usb_mass_storage.c | 25 ++++++++++++++-----------
 1 file changed, 14 insertions(+), 11 deletions(-)

diff --git a/common/cmd_usb_mass_storage.c b/common/cmd_usb_mass_storage.c
index 5ec5da6d4f04..b08f1e5be141 100644
--- a/common/cmd_usb_mass_storage.c
+++ b/common/cmd_usb_mass_storage.c
@@ -9,7 +9,6 @@
 #include <common.h>
 #include <command.h>
 #include <g_dnl.h>
-#include <mmc.h>
 #include <part.h>
 #include <usb.h>
 #include <usb_mass_storage.h>
@@ -40,17 +39,22 @@ static struct ums ums_dev = {
 	.name = "UMS disk",
 };
 
-struct ums *ums_init(unsigned int dev_num)
+struct ums *ums_init(const char *devtype, const char *devnum)
 {
-	struct mmc *mmc = NULL;
+	block_dev_desc_t *block_dev;
+	int ret;
 
-	mmc = find_mmc_device(dev_num);
-	if (!mmc)
+	ret = get_device(devtype, devnum, &block_dev);
+	if (ret < 0)
 		return NULL;
 
-	ums_dev.block_dev = &mmc->block_dev;
+	/* f_mass_storage.c assumes SECTOR_SIZE sectors */
+	if (block_dev->blksz != SECTOR_SIZE)
+		return NULL;
+
+	ums_dev.block_dev = block_dev;
 	ums_dev.start_sector = 0;
-	ums_dev.num_sectors = mmc->capacity / SECTOR_SIZE;
+	ums_dev.num_sectors = block_dev->lba;
 
 	printf("UMS: disk start sector: %#x, count: %#x\n",
 	       ums_dev.start_sector, ums_dev.num_sectors);
@@ -65,11 +69,10 @@ int do_usb_mass_storage(cmd_tbl_t *cmdtp, int flag,
 		return CMD_RET_USAGE;
 
 	const char *usb_controller = argv[1];
-	const char *mmc_devstring  = argv[2];
-
-	unsigned int dev_num = simple_strtoul(mmc_devstring, NULL, 0);
+	const char *devtype = "mmc";
+	const char *devnum  = argv[2];
 
-	struct ums *ums = ums_init(dev_num);
+	struct ums *ums = ums_init(devtype, devnum);
 	if (!ums)
 		return CMD_RET_FAILURE;
 
-- 
1.8.1.5

[U-Boot] [PATCH 7/8] ums: move all variable declarations to the start of the block

[Stephen Warren]

From: Stephen Warren <swarren@nvidia.com>

It's easier to assign values to the variables inside an if statement body
if the assignment and declaration are separate.

Cc: Przemyslaw Marczak <p.marczak@samsung.com>
Cc: Lukasz Majewski <l.majewski@samsung.com>
Signed-off-by: Stephen Warren <swarren@nvidia.com>
---
 common/cmd_usb_mass_storage.c | 24 ++++++++++++++++--------
 1 file changed, 16 insertions(+), 8 deletions(-)

diff --git a/common/cmd_usb_mass_storage.c b/common/cmd_usb_mass_storage.c
index b08f1e5be141..31dd37c7c66f 100644
--- a/common/cmd_usb_mass_storage.c
+++ b/common/cmd_usb_mass_storage.c
@@ -65,25 +65,33 @@ struct ums *ums_init(const char *devtype, const char *devnum)
 int do_usb_mass_storage(cmd_tbl_t *cmdtp, int flag,
 			       int argc, char * const argv[])
 {
+	const char *usb_controller;
+	const char *devtype;
+	const char *devnum;
+	struct ums *ums;
+	unsigned int controller_index;
+	int rc;
+	int cable_ready_timeout __maybe_unused;
+
 	if (argc < 3)
 		return CMD_RET_USAGE;
 
-	const char *usb_controller = argv[1];
-	const char *devtype = "mmc";
-	const char *devnum  = argv[2];
+	usb_controller = argv[1];
+	devtype = "mmc";
+	devnum  = argv[2];
 
-	struct ums *ums = ums_init(devtype, devnum);
+	ums = ums_init(devtype, devnum);
 	if (!ums)
 		return CMD_RET_FAILURE;
 
-	unsigned int controller_index = (unsigned int)(simple_strtoul(
-					usb_controller,	NULL, 0));
+	controller_index = (unsigned int)(simple_strtoul(
+				usb_controller,	NULL, 0));
 	if (board_usb_init(controller_index, USB_INIT_DEVICE)) {
 		error("Couldn't init USB controller.");
 		return CMD_RET_FAILURE;
 	}
 
-	int rc = fsg_init(ums);
+	rc = fsg_init(ums);
 	if (rc) {
 		error("fsg_init failed");
 		return CMD_RET_FAILURE;
@@ -93,7 +101,7 @@ int do_usb_mass_storage(cmd_tbl_t *cmdtp, int flag,
 
 #ifdef CONFIG_USB_CABLE_CHECK
 	/* Timeout unit: seconds */
-	int cable_ready_timeout = UMS_CABLE_READY_TIMEOUT;
+	cable_ready_timeout = UMS_CABLE_READY_TIMEOUT;
 
 	if (!usb_cable_connected()) {
 		puts("Please connect USB cable.\n");
-- 
1.8.1.5

[U-Boot] [PATCH 7/8] ums: move all variable declarations to the start of the block

[Marek Vasut]

On Wednesday, April 30, 2014 at 11:13:21 PM, Stephen Warren wrote:
> From: Stephen Warren <swarren@nvidia.com>
> 
> It's easier to assign values to the variables inside an if statement body
> if the assignment and declaration are separate.
> 
> Cc: Przemyslaw Marczak <p.marczak@samsung.com>
> Cc: Lukasz Majewski <l.majewski@samsung.com>
> Signed-off-by: Stephen Warren <swarren@nvidia.com>
> ---
>  common/cmd_usb_mass_storage.c | 24 ++++++++++++++++--------
>  1 file changed, 16 insertions(+), 8 deletions(-)

This doesn't apply on u-boot-usb/master, sorry.

Best regards,
Marek Vasut

[U-Boot] [PATCH 7/8] ums: move all variable declarations to the start of the block

[Stephen Warren]

On 05/01/2014 04:51 AM, Marek Vasut wrote:
> On Wednesday, April 30, 2014 at 11:13:21 PM, Stephen Warren wrote:
>> From: Stephen Warren <swarren@nvidia.com>
>>
>> It's easier to assign values to the variables inside an if statement body
>> if the assignment and declaration are separate.
>>
>> Cc: Przemyslaw Marczak <p.marczak@samsung.com>
>> Cc: Lukasz Majewski <l.majewski@samsung.com>
>> Signed-off-by: Stephen Warren <swarren@nvidia.com>
>> ---
>>  common/cmd_usb_mass_storage.c | 24 ++++++++++++++++--------
>>  1 file changed, 16 insertions(+), 8 deletions(-)
> 
> This doesn't apply on u-boot-usb/master, sorry.

Is this the first patch in the series which doesn't apply? Did you apply
any of the earlier patches? I guess I'll go try to rebase the series and
find out myself...

This is why we need a "uboot-next" just like linux-next, and stricter
controls on which git repos take patches that touch subsystem code, to
avoid cross-repo conflicts:-(

[U-Boot] [PATCH 7/8] ums: move all variable declarations to the start of the block

[Lukasz Majewski]

Hi Stephen,

> 
> On 05/01/2014 04:51 AM, Marek Vasut wrote:
> > On Wednesday, April 30, 2014 at 11:13:21 PM, Stephen Warren wrote:
> >> From: Stephen Warren <swarren@nvidia.com>
> >>
> >> It's easier to assign values to the variables inside an if
> >> statement body if the assignment and declaration are separate.
> >>
> >> Cc: Przemyslaw Marczak <p.marczak@samsung.com>
> >> Cc: Lukasz Majewski <l.majewski@samsung.com>
> >> Signed-off-by: Stephen Warren <swarren@nvidia.com>
> >> ---
> >>  common/cmd_usb_mass_storage.c | 24 ++++++++++++++++--------
> >>  1 file changed, 16 insertions(+), 8 deletions(-)
> > 
> > This doesn't apply on u-boot-usb/master, sorry.
> 
> Is this the first patch in the series which doesn't apply? Did you
> apply any of the earlier patches? I guess I'll go try to rebase the
> series and find out myself...
> 
> This is why we need a "uboot-next" just like linux-next, and stricter
> controls on which git repos take patches that touch subsystem code, to
> avoid cross-repo conflicts:-(

+1

Yesterday Marek and I agreed that some patches regarding DFU and
gadgets (patches 1 to 8) done by Mateusz:

http://patchwork.ozlabs.org/patch/343517/

will be go through u-boot-dfu tree, which will be rebased on top of
u-boot-usb tree. Afterwards I will send PR to Marek.

Unfortunately I didn't manage to add them to u-boot-denx tree yesterday.
I will do it on Monday as well as the review regarding your ums work. I
hope, that you can wait until then.


Best regards,
Lukasz Majewski

> _______________________________________________
> U-Boot mailing list
> U-Boot at lists.denx.de
> http://lists.denx.de/mailman/listinfo/u-boot

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20140501/506ce407/attachment.pgp>

[U-Boot] [PATCH 7/8] ums: move all variable declarations to the start of the block

[Stephen Warren]

On 05/01/2014 01:43 PM, Lukasz Majewski wrote:
> Hi Stephen,
> 
>>
>> On 05/01/2014 04:51 AM, Marek Vasut wrote:
>>> On Wednesday, April 30, 2014 at 11:13:21 PM, Stephen Warren wrote:
>>>> From: Stephen Warren <swarren@nvidia.com>
>>>>
>>>> It's easier to assign values to the variables inside an if
>>>> statement body if the assignment and declaration are separate.
>>>>
>>>> Cc: Przemyslaw Marczak <p.marczak@samsung.com>
>>>> Cc: Lukasz Majewski <l.majewski@samsung.com>
>>>> Signed-off-by: Stephen Warren <swarren@nvidia.com>
>>>> ---
>>>>  common/cmd_usb_mass_storage.c | 24 ++++++++++++++++--------
>>>>  1 file changed, 16 insertions(+), 8 deletions(-)
>>>
>>> This doesn't apply on u-boot-usb/master, sorry.
>>
>> Is this the first patch in the series which doesn't apply? Did you
>> apply any of the earlier patches? I guess I'll go try to rebase the
>> series and find out myself...
>>
>> This is why we need a "uboot-next" just like linux-next, and stricter
>> controls on which git repos take patches that touch subsystem code, to
>> avoid cross-repo conflicts:-(
> 
> +1
> 
> Yesterday Marek and I agreed that some patches regarding DFU and
> gadgets (patches 1 to 8) done by Mateusz:
> 
> http://patchwork.ozlabs.org/patch/343517/
> 
> will be go through u-boot-dfu tree, which will be rebased on top of
> u-boot-usb tree. Afterwards I will send PR to Marek.
> 
> Unfortunately I didn't manage to add them to u-boot-denx tree yesterday.
> I will do it on Monday as well as the review regarding your ums work. I
> hope, that you can wait until then.

OK, that's fine.

FWIW, I have applied that series locally and rebased my patches on top
of it. If you want to see the latest version, feel free to look at:

git://github.com/swarren/u-boot.git tegra_dev

Note that branch gets rebased as I do local development.

[U-Boot] [PATCH 7/8] ums: move all variable declarations to the start of the block

[Przemyslaw Marczak]

Hello Stephen,

On 05/01/2014 11:47 PM, Stephen Warren wrote:
> On 05/01/2014 01:43 PM, Lukasz Majewski wrote:
>> Hi Stephen,
>>
>>>
>>> On 05/01/2014 04:51 AM, Marek Vasut wrote:
>>>> On Wednesday, April 30, 2014 at 11:13:21 PM, Stephen Warren wrote:
>>>>> From: Stephen Warren <swarren@nvidia.com>
>>>>>
>>>>> It's easier to assign values to the variables inside an if
>>>>> statement body if the assignment and declaration are separate.
>>>>>
>>>>> Cc: Przemyslaw Marczak <p.marczak@samsung.com>
>>>>> Cc: Lukasz Majewski <l.majewski@samsung.com>
>>>>> Signed-off-by: Stephen Warren <swarren@nvidia.com>
>>>>> ---
>>>>>   common/cmd_usb_mass_storage.c | 24 ++++++++++++++++--------
>>>>>   1 file changed, 16 insertions(+), 8 deletions(-)
>>>>
>>>> This doesn't apply on u-boot-usb/master, sorry.
>>>
>>> Is this the first patch in the series which doesn't apply? Did you
>>> apply any of the earlier patches? I guess I'll go try to rebase the
>>> series and find out myself...
>>>
>>> This is why we need a "uboot-next" just like linux-next, and stricter
>>> controls on which git repos take patches that touch subsystem code, to
>>> avoid cross-repo conflicts:-(
>>
>> +1
>>
>> Yesterday Marek and I agreed that some patches regarding DFU and
>> gadgets (patches 1 to 8) done by Mateusz:
>>
>> http://patchwork.ozlabs.org/patch/343517/
>>
>> will be go through u-boot-dfu tree, which will be rebased on top of
>> u-boot-usb tree. Afterwards I will send PR to Marek.
>>
>> Unfortunately I didn't manage to add them to u-boot-denx tree yesterday.
>> I will do it on Monday as well as the review regarding your ums work. I
>> hope, that you can wait until then.
>
> OK, that's fine.
>
> FWIW, I have applied that series locally and rebased my patches on top
> of it. If you want to see the latest version, feel free to look at:
>
> git://github.com/swarren/u-boot.git tegra_dev
>
> Note that branch gets rebased as I do local development.
>

I tested your tegra_dev branch on trats2 and also on goni(after apply 
the rest of Mateusz Zalega patches) and it works fine.

So for this patch set:
Acked-by: Przemyslaw Marczak <p.marczak@samsung.com>

Thanks,
-- 
Przemyslaw Marczak
Samsung R&D Institute Poland
Samsung Electronics
p.marczak at samsung.com

[U-Boot] [PATCH 7/8] ums: move all variable declarations to the start of the block

[Stephen Warren]

On 05/05/2014 03:16 AM, Przemyslaw Marczak wrote:
...
>>>>> On Wednesday, April 30, 2014 at 11:13:21 PM, Stephen Warren wrote:
>>>>>> It's easier to assign values to the variables inside an if
>>>>>> statement body if the assignment and declaration are separate.
...
>> FWIW, I have applied that series locally and rebased my patches on top
>> of it. If you want to see the latest version, feel free to look at:
>>
>> git://github.com/swarren/u-boot.git tegra_dev
...
> I tested your tegra_dev branch on trats2 and also on goni(after apply
> the rest of Mateusz Zalega patches) and it works fine.
> 
> So for this patch set:
> Acked-by: Przemyslaw Marczak <p.marczak@samsung.com>

Good new:-) Thanks for the testing.

[U-Boot] [PATCH 7/8] ums: move all variable declarations to the start of the block

[Marek Vasut]

On Thursday, May 01, 2014 at 06:58:50 PM, Stephen Warren wrote:
> On 05/01/2014 04:51 AM, Marek Vasut wrote:
> > On Wednesday, April 30, 2014 at 11:13:21 PM, Stephen Warren wrote:
> >> From: Stephen Warren <swarren@nvidia.com>
> >> 
> >> It's easier to assign values to the variables inside an if statement
> >> body if the assignment and declaration are separate.
> >> 
> >> Cc: Przemyslaw Marczak <p.marczak@samsung.com>
> >> Cc: Lukasz Majewski <l.majewski@samsung.com>
> >> Signed-off-by: Stephen Warren <swarren@nvidia.com>
> >> ---
> >> 
> >>  common/cmd_usb_mass_storage.c | 24 ++++++++++++++++--------
> >>  1 file changed, 16 insertions(+), 8 deletions(-)
> > 
> > This doesn't apply on u-boot-usb/master, sorry.
> 
> Is this the first patch in the series which doesn't apply? Did you apply
> any of the earlier patches? I guess I'll go try to rebase the series and
> find out myself...

7/8 doesn't apply .

> This is why we need a "uboot-next" just like linux-next, and stricter
> controls on which git repos take patches that touch subsystem code, to
> avoid cross-repo conflicts:-(

Yep, u-boot-next would be nice and so would be u-boot-stable . Volunteers 
welcome ;-)

Best regards,
Marek Vasut

[U-Boot] [PATCH 8/8] ums: allow the user to specify the device type

[Stephen Warren]

From: Stephen Warren <swarren@nvidia.com>

Allow an optional devtype parameter to the ums command, which specifies
the type of the device to be exported. This could allow exporting a SATA
or even another USB device.

Cc: Przemyslaw Marczak <p.marczak@samsung.com>
Cc: Lukasz Majewski <l.majewski@samsung.com>
Signed-off-by: Stephen Warren <swarren@nvidia.com>
---
 common/cmd_usb_mass_storage.c | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

diff --git a/common/cmd_usb_mass_storage.c b/common/cmd_usb_mass_storage.c
index 31dd37c7c66f..2ef5c0912e9c 100644
--- a/common/cmd_usb_mass_storage.c
+++ b/common/cmd_usb_mass_storage.c
@@ -77,8 +77,13 @@ int do_usb_mass_storage(cmd_tbl_t *cmdtp, int flag,
 		return CMD_RET_USAGE;
 
 	usb_controller = argv[1];
-	devtype = "mmc";
-	devnum  = argv[2];
+	if (argc >= 4) {
+		devtype = argv[2];
+		devnum  = argv[3];
+	} else {
+		devtype = "mmc";
+		devnum  = argv[2];
+	}
 
 	ums = ums_init(devtype, devnum);
 	if (!ums)
@@ -146,7 +151,8 @@ exit:
 	return CMD_RET_SUCCESS;
 }
 
-U_BOOT_CMD(ums, CONFIG_SYS_MAXARGS, 1, do_usb_mass_storage,
+U_BOOT_CMD(ums, 4, 1, do_usb_mass_storage,
 	"Use the UMS [User Mass Storage]",
-	"ums <USB_controller> <mmc_dev>  e.g. ums 0 0"
+	"ums <USB_controller> [<devtype>] <devnum>  e.g. ums 0 mmc 0\n"
+	"    devtype defaults to mmc"
 );
-- 
1.8.1.5