[U-Boot] [PATCH 00/10] x86: Early debug enhancements

[U-Boot] [PATCH 00/10] x86: Early debug enhancements

[Simon Glass]

This series adds a few more features to the debug UART, making it possible
to use it even when the board needs to perform additional init for the UART
to work. The debug UART is anabled on link.

Some minor PCI tidy-ups are also included.

Link currently does not boot due to lack of early malloc() memory. For now
just increase the size.


Simon Glass (10):
  x86: chromebook_link: Expand early malloc() memory
  malloc_simple: Add debug() information
  dm: pci: Tidy up auto-config error handling
  dm: pci: Correct a few debug() statements
  dm: pci: Adjust pci_find_and_bind_driver() to return -EPERM
  debug_uart: Adjust the declaration of debug_uart_init()
  debug_uart: Support board-specific UART initialisation
  debug_uart: Add an option to announce the debug UART
  x86: Init the debug UART if enabled
  x86: chromebook_link: Enable the debug UART

 arch/x86/cpu/ivybridge/cpu.c      |  7 ++++++
 arch/x86/cpu/start.S              |  3 +++
 common/malloc_simple.c            |  4 ++++
 configs/chromebook_link_defconfig | 11 ++++++---
 drivers/pci/pci-uclass.c          | 50 ++++++++++++++++++++++++++-------------
 drivers/serial/Kconfig            | 21 ++++++++++++++++
 drivers/serial/ns16550.c          |  2 +-
 drivers/serial/serial_efi.c       |  2 +-
 drivers/serial/serial_s5p.c       |  2 +-
 include/debug_uart.h              | 30 +++++++++++++++++++++--
 lib/efi/efi_stub.c                |  2 +-
 11 files changed, 108 insertions(+), 26 deletions(-)

-- 
2.6.0.rc0.131.gf624c3d

[U-Boot] [PATCH 01/10] x86: chromebook_link: Expand early malloc() memory

[Simon Glass]

Now that PCI bridges are probed before relocation we need additional memory.
Each PCI bridge takes 240 bytes at present since it uses the same uclass as
the PCI controller. Probably we should split this out so that bridges have
their own uclass.

Expand the memory on link so that it works correctly.

Signed-off-by: Simon Glass <sjg@chromium.org>
---

 configs/chromebook_link_defconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/configs/chromebook_link_defconfig b/configs/chromebook_link_defconfig
index 9855736..b3ae925 100644
--- a/configs/chromebook_link_defconfig
+++ b/configs/chromebook_link_defconfig
@@ -1,4 +1,5 @@
 CONFIG_X86=y
+CONFIG_SYS_MALLOC_F_LEN=0x1800
 CONFIG_VENDOR_GOOGLE=y
 CONFIG_DEFAULT_DEVICE_TREE="chromebook_link"
 CONFIG_TARGET_CHROMEBOOK_LINK=y
-- 
2.6.0.rc0.131.gf624c3d

[U-Boot] [PATCH 02/10] malloc_simple: Add debug() information

[Simon Glass]

It's useful to get a a trace of memory allocations in early init. Add a
debug() call to provide that. It can be enabled by adding '#define DEBUG'
to the top of the file.

Signed-off-by: Simon Glass <sjg@chromium.org>
---

 common/malloc_simple.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/common/malloc_simple.c b/common/malloc_simple.c
index 134e059..e357827 100644
--- a/common/malloc_simple.c
+++ b/common/malloc_simple.c
@@ -19,10 +19,13 @@ void *malloc_simple(size_t bytes)
 	void *ptr;
 
 	new_ptr = gd->malloc_ptr + bytes;
+	debug("%s: size=%zx, ptr=%lx, limit=%lx\n", __func__, bytes, new_ptr,
+	      gd->malloc_limit);
 	if (new_ptr > gd->malloc_limit)
 		return NULL;
 	ptr = map_sysmem(gd->malloc_base + gd->malloc_ptr, bytes);
 	gd->malloc_ptr = ALIGN(new_ptr, sizeof(new_ptr));
+
 	return ptr;
 }
 
@@ -37,6 +40,7 @@ void *memalign_simple(size_t align, size_t bytes)
 		return NULL;
 	ptr = map_sysmem(addr, bytes);
 	gd->malloc_ptr = ALIGN(new_ptr, sizeof(new_ptr));
+
 	return ptr;
 }
 
-- 
2.6.0.rc0.131.gf624c3d

[U-Boot] [PATCH 03/10] dm: pci: Tidy up auto-config error handling

[Simon Glass]

When the auto-configuration process fails for a device (generally due to
lack of memory) we should return the error correctly so that we don't
continue to try memory allocations which will fail.

Adjust the code to check for errors and abort if something goes wrong.

Signed-off-by: Simon Glass <sjg@chromium.org>
---

 drivers/pci/pci-uclass.c | 15 ++++++++++++---
 1 file changed, 12 insertions(+), 3 deletions(-)

diff --git a/drivers/pci/pci-uclass.c b/drivers/pci/pci-uclass.c
index b25298f..437c81a 100644
--- a/drivers/pci/pci-uclass.c
+++ b/drivers/pci/pci-uclass.c
@@ -381,9 +381,13 @@ int pci_auto_config_devices(struct udevice *bus)
 	     !ret && dev;
 	     ret = device_find_next_child(&dev)) {
 		unsigned int max_bus;
+		int ret;
 
 		debug("%s: device %s\n", __func__, dev->name);
-		max_bus = pciauto_config_device(hose, pci_get_bdf(dev));
+		ret = pciauto_config_device(hose, pci_get_bdf(dev));
+		if (ret < 0)
+			return ret;
+		max_bus = ret;
 		sub_bus = max(sub_bus, max_bus);
 	}
 	debug("%s: done\n", __func__);
@@ -757,6 +761,8 @@ static int pci_uclass_post_probe(struct udevice *bus)
 
 #ifdef CONFIG_PCI_PNP
 	ret = pci_auto_config_devices(bus);
+	if (ret < 0)
+		return ret;
 #endif
 
 #if defined(CONFIG_X86) && defined(CONFIG_HAVE_FSP)
@@ -773,11 +779,14 @@ static int pci_uclass_post_probe(struct udevice *bus)
 	 * Note we only call this 1) after U-Boot is relocated, and 2)
 	 * root bus has finished probing.
 	 */
-	if ((gd->flags & GD_FLG_RELOC) && (bus->seq == 0))
+	if ((gd->flags & GD_FLG_RELOC) && (bus->seq == 0)) {
 		ret = fsp_init_phase_pci();
+		if (ret)
+			return ret;
+	}
 #endif
 
-	return ret < 0 ? ret : 0;
+	return 0;
 }
 
 static int pci_uclass_child_post_bind(struct udevice *dev)
-- 
2.6.0.rc0.131.gf624c3d

[U-Boot] [PATCH 04/10] dm: pci: Correct a few debug() statements

[Simon Glass]

One debug() statement is missing a newline. The other has a repeated word.
Fix these.

Signed-off-by: Simon Glass <sjg@chromium.org>
---

 drivers/pci/pci-uclass.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/pci/pci-uclass.c b/drivers/pci/pci-uclass.c
index 437c81a..7347160 100644
--- a/drivers/pci/pci-uclass.c
+++ b/drivers/pci/pci-uclass.c
@@ -418,7 +418,7 @@ int dm_pci_hose_probe_bus(struct pci_controller *hose, pci_dev_t bdf)
 
 	ret = device_probe(bus);
 	if (ret) {
-		debug("%s: Cannot probe bus bus %s: %d\n", __func__, bus->name,
+		debug("%s: Cannot probe bus %s: %d\n", __func__, bus->name,
 		      ret);
 		return ret;
 	}
@@ -537,7 +537,7 @@ static int pci_find_and_bind_driver(struct udevice *parent,
 
 	ret = device_bind_driver(parent, drv, str, devp);
 	if (ret) {
-		debug("%s: Failed to bind generic driver: %d", __func__, ret);
+		debug("%s: Failed to bind generic driver: %d\n", __func__, ret);
 		return ret;
 	}
 	debug("%s: No match found: bound generic driver instead\n", __func__);
-- 
2.6.0.rc0.131.gf624c3d

[U-Boot] [PATCH 05/10] dm: pci: Adjust pci_find_and_bind_driver() to return -EPERM

[Simon Glass]

The current code returns 0 even if it failed to find or bind a driver. The
caller then has to check the returned device to see if it is NULL. It is
better to return an error code in this case so that it is clear what
happened.

Adjust the code to return -EPERM, indicating that the device was not bound
because it is not needed for pre-relocation use. Add comments so that the
return value is clear.

Signed-off-by: Simon Glass <sjg@chromium.org>
---

 drivers/pci/pci-uclass.c | 31 +++++++++++++++++++------------
 1 file changed, 19 insertions(+), 12 deletions(-)

diff --git a/drivers/pci/pci-uclass.c b/drivers/pci/pci-uclass.c
index 7347160..54285ee 100644
--- a/drivers/pci/pci-uclass.c
+++ b/drivers/pci/pci-uclass.c
@@ -458,10 +458,17 @@ static bool pci_match_one_id(const struct pci_device_id *id,
  * pci_find_and_bind_driver() - Find and bind the right PCI driver
  *
  * This only looks at certain fields in the descriptor.
+ *
+ * @parent:	Parent bus
+ * @find_id:	Specification of the driver to find
+ * @bdf:	Bus/device/function addreess - see PCI_BDF()
+ * @devp:	Returns a pointer to the device created
+ * @return 0 if OK, -EPERM if the device is not needed before relocation and
+ *	   therefore was not created, other -ve value on error
  */
 static int pci_find_and_bind_driver(struct udevice *parent,
-				    struct pci_device_id *find_id, pci_dev_t bdf,
-				    struct udevice **devp)
+				    struct pci_device_id *find_id,
+				    pci_dev_t bdf, struct udevice **devp)
 {
 	struct pci_driver_entry *start, *entry;
 	const char *drv;
@@ -497,7 +504,7 @@ static int pci_find_and_bind_driver(struct udevice *parent,
 			 */
 			if (!(gd->flags & GD_FLG_RELOC) &&
 			    !(drv->flags & DM_FLAG_PRE_RELOC))
-				return 0;
+				return -EPERM;
 
 			/*
 			 * We could pass the descriptor to the driver as
@@ -525,7 +532,7 @@ static int pci_find_and_bind_driver(struct udevice *parent,
 	 * limited (ie: using Cache As RAM).
 	 */
 	if (!(gd->flags & GD_FLG_RELOC) && !bridge)
-		return 0;
+		return -EPERM;
 
 	/* Bind a generic driver so that the device can be used */
 	sprintf(name, "pci_%x:%x.%x", parent->seq, PCI_DEV(bdf),
@@ -613,17 +620,17 @@ int pci_bind_bus_devices(struct udevice *bus)
 			ret = pci_find_and_bind_driver(bus, &find_id, bdf,
 						       &dev);
 		}
-		if (ret)
+		if (ret == -EPERM)
+			continue;
+		else if (ret)
 			return ret;
 
 		/* Update the platform data */
-		if (dev) {
-			pplat = dev_get_parent_platdata(dev);
-			pplat->devfn = PCI_MASK_BUS(bdf);
-			pplat->vendor = vendor;
-			pplat->device = device;
-			pplat->class = class;
-		}
+		pplat = dev_get_parent_platdata(dev);
+		pplat->devfn = PCI_MASK_BUS(bdf);
+		pplat->vendor = vendor;
+		pplat->device = device;
+		pplat->class = class;
 	}
 
 	return 0;
-- 
2.6.0.rc0.131.gf624c3d

[U-Boot] [PATCH 06/10] debug_uart: Adjust the declaration of debug_uart_init()

[Simon Glass]

We want to be able to add other common code to this function. So change the
driver's version to have an underscore before it, just like
_debug_uart_putc(). Define debug_uart_init() to call this version.

Update all drivers to this new method.

Signed-off-by: Simon Glass <sjg@chromium.org>
---

 drivers/serial/ns16550.c    | 2 +-
 drivers/serial/serial_efi.c | 2 +-
 drivers/serial/serial_s5p.c | 2 +-
 include/debug_uart.h        | 9 +++++++--
 lib/efi/efi_stub.c          | 2 +-
 5 files changed, 11 insertions(+), 6 deletions(-)

diff --git a/drivers/serial/ns16550.c b/drivers/serial/ns16550.c
index 6275a11..6433844 100644
--- a/drivers/serial/ns16550.c
+++ b/drivers/serial/ns16550.c
@@ -257,7 +257,7 @@ int NS16550_tstc(NS16550_t com_port)
 			(1 << CONFIG_DEBUG_UART_SHIFT), \
 		CONFIG_DEBUG_UART_SHIFT)
 
-void debug_uart_init(void)
+static inline void _debug_uart_init(void)
 {
 	struct NS16550 *com_port = (struct NS16550 *)CONFIG_DEBUG_UART_BASE;
 	int baud_divisor;
diff --git a/drivers/serial/serial_efi.c b/drivers/serial/serial_efi.c
index cf57d89..ee3029b 100644
--- a/drivers/serial/serial_efi.c
+++ b/drivers/serial/serial_efi.c
@@ -107,7 +107,7 @@ static int serial_efi_pending(struct udevice *dev, bool input)
  * There is nothing to init here since the EFI console is already running by
  * the time we enter U-Boot.
  */
-void debug_uart_init(void)
+void _debug_uart_init(void)
 {
 }
 
diff --git a/drivers/serial/serial_s5p.c b/drivers/serial/serial_s5p.c
index 3f0b588..e7adf0d 100644
--- a/drivers/serial/serial_s5p.c
+++ b/drivers/serial/serial_s5p.c
@@ -207,7 +207,7 @@ U_BOOT_DRIVER(serial_s5p) = {
 
 #include <debug_uart.h>
 
-void debug_uart_init(void)
+void _debug_uart_init(void)
 {
 	struct s5p_uart *uart = (struct s5p_uart *)CONFIG_DEBUG_UART_BASE;
 
diff --git a/include/debug_uart.h b/include/debug_uart.h
index a75e377..257ba00 100644
--- a/include/debug_uart.h
+++ b/include/debug_uart.h
@@ -38,7 +38,7 @@
  * To enable the debug UART in your serial driver:
  *
  * - #include <debug_uart.h>
- * - Define debug_uart_init(), trying to avoid using the stack
+ * - Define _debug_uart_init(), trying to avoid using the stack
  * - Define _debug_uart_putc() as static inline (avoiding stack usage)
  * - Immediately afterwards, add DEBUG_UART_FUNCS to define the rest of the
  *     functionality (printch(), etc.)
@@ -132,6 +132,11 @@ void printhex8(uint value);
 	void printhex8(uint value) \
 	{ \
 		printhex(value, 8); \
-	}
+	} \
+\
+	void debug_uart_init(void) \
+	{ \
+		_debug_uart_init(); \
+	} \
 
 #endif
diff --git a/lib/efi/efi_stub.c b/lib/efi/efi_stub.c
index d4d3e49..e138709 100644
--- a/lib/efi/efi_stub.c
+++ b/lib/efi/efi_stub.c
@@ -59,7 +59,7 @@ struct __packed desctab_info {
  * considering if we start needing more U-Boot functionality. Note that we
  * could then move get_codeseg32() to arch/x86/cpu/cpu.c.
  */
-void debug_uart_init(void)
+void _debug_uart_init(void)
 {
 }
 
-- 
2.6.0.rc0.131.gf624c3d

[U-Boot] [PATCH 07/10] debug_uart: Support board-specific UART initialisation

[Simon Glass]

Some boards need to set things up before the debug UART can be used. On
these boards a call to debug_uart_init() is insufficient. When this option
is enabled, the function board_debug_uart_init() will be called when
debug_uart_init() is called. You can put any code here that is needed to
set up the UART ready for use, such as set pin multiplexing or enable
clocks.

Signed-off-by: Simon Glass <sjg@chromium.org>
---

 drivers/serial/Kconfig | 11 +++++++++++
 include/debug_uart.h   | 14 ++++++++++++++
 2 files changed, 25 insertions(+)

diff --git a/drivers/serial/Kconfig b/drivers/serial/Kconfig
index ccb80d2..4418a57 100644
--- a/drivers/serial/Kconfig
+++ b/drivers/serial/Kconfig
@@ -109,6 +109,17 @@ config DEBUG_UART_SHIFT
 	  value. Use this value to specify the shift to use, where 0=byte
 	  registers, 2=32-bit word registers, etc.
 
+config DEBUG_UART_BOARD_INIT
+	bool "Enable board-specific debug UART init"
+	depends on DEBUG_UART
+	help
+	  Some boards need to set things up before the debug UART can be used.
+	  On these boards a call to debug_uart_init() is insufficient. When
+	  this option is enabled, the function board_debug_uart_init() will
+	  be called when debug_uart_init() is called. You can put any code
+	  here that is needed to set up the UART ready for use, such as set
+	  pin multiplexing or enable clocks.
+
 config ROCKCHIP_SERIAL
 	bool "Rockchip on-chip UART support"
 	depends on ARCH_UNIPHIER && DM_SERIAL
diff --git a/include/debug_uart.h b/include/debug_uart.h
index 257ba00..a6b7ce8 100644
--- a/include/debug_uart.h
+++ b/include/debug_uart.h
@@ -42,6 +42,11 @@
  * - Define _debug_uart_putc() as static inline (avoiding stack usage)
  * - Immediately afterwards, add DEBUG_UART_FUNCS to define the rest of the
  *     functionality (printch(), etc.)
+ *
+ * If your board needs additional init for the UART to work, enable
+ * CONFIG_DEBUG_UART_BOARD_INIT and write a function called
+ * board_debug_uart_init() to perform that init. When debug_uart_init() is
+ * called, the init will happen automatically.
  */
 
 /**
@@ -57,6 +62,14 @@
  */
 void debug_uart_init(void);
 
+#ifdef CONFIG_DEBUG_UART_BOARD_INIT
+void board_debug_uart_init(void);
+#else
+static inline void board_debug_uart_init(void)
+{
+}
+#endif
+
 /**
  * printch() - Output a character to the debug UART
  *
@@ -136,6 +149,7 @@ void printhex8(uint value);
 \
 	void debug_uart_init(void) \
 	{ \
+		board_debug_uart_init(); \
 		_debug_uart_init(); \
 	} \
 
-- 
2.6.0.rc0.131.gf624c3d

[U-Boot] [PATCH 08/10] debug_uart: Add an option to announce the debug UART

[Simon Glass]

It is useful to see a message from the debug UART early during boot so that
you know things are working. Add an option to enable this. The message will
be displayed as soon as debug_uart_init() is called.

Signed-off-by: Simon Glass <sjg@chromium.org>
---

 drivers/serial/Kconfig | 10 ++++++++++
 include/debug_uart.h   |  7 +++++++
 2 files changed, 17 insertions(+)

diff --git a/drivers/serial/Kconfig b/drivers/serial/Kconfig
index 4418a57..fa03d17 100644
--- a/drivers/serial/Kconfig
+++ b/drivers/serial/Kconfig
@@ -120,6 +120,16 @@ config DEBUG_UART_BOARD_INIT
 	  here that is needed to set up the UART ready for use, such as set
 	  pin multiplexing or enable clocks.
 
+config DEBUG_UART_ANNOUNCE
+	bool "Show a message when the debug UART starts up"
+	depends on DEBUG_UART
+	help
+	  Enable this option to show a message when the debug UART is ready
+	  for use. You will see a message like "<debug_uart> " as soon as
+	  U-Boot has the UART ready for use (i.e. your code calls
+	  debug_uart_init()). This can be useful just as a check that
+	  everything is working.
+
 config ROCKCHIP_SERIAL
 	bool "Rockchip on-chip UART support"
 	depends on ARCH_UNIPHIER && DM_SERIAL
diff --git a/include/debug_uart.h b/include/debug_uart.h
index a6b7ce8..5d5349b 100644
--- a/include/debug_uart.h
+++ b/include/debug_uart.h
@@ -105,6 +105,12 @@ void printhex4(uint value);
  */
 void printhex8(uint value);
 
+#ifdef CONFIG_DEBUG_UART_ANNOUNCE
+#define _DEBUG_UART_ANNOUNCE	printascii("<debug_uart> ");
+#else
+#define _DEBUG_UART_ANNOUNCE
+#endif
+
 /*
  * Now define some functions - this should be inserted into the serial driver
  */
@@ -151,6 +157,7 @@ void printhex8(uint value);
 	{ \
 		board_debug_uart_init(); \
 		_debug_uart_init(); \
+		_DEBUG_UART_ANNOUNCE \
 	} \
 
 #endif
-- 
2.6.0.rc0.131.gf624c3d

[U-Boot] [PATCH 09/10] x86: Init the debug UART if enabled

[Simon Glass]

If the debug UART is enabled, get it ready for use at the earliest possible
opportunity. This is not actually very early, but until we have a stack it
is difficult to make it work.

Signed-off-by: Simon Glass <sjg@chromium.org>
---

 arch/x86/cpu/start.S | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/arch/x86/cpu/start.S b/arch/x86/cpu/start.S
index d072825..e6e9cb5 100644
--- a/arch/x86/cpu/start.S
+++ b/arch/x86/cpu/start.S
@@ -134,6 +134,9 @@ car_init_ret:
 	 * mov   $'a', %eax
 	 * call  printch
 	 */
+#ifdef CONFIG_DEBUG_UART
+	call  debug_uart_init
+#endif
 
 	/* Get address of global_data */
 	mov	%fs:0, %edx
-- 
2.6.0.rc0.131.gf624c3d

[U-Boot] [PATCH 10/10] x86: chromebook_link: Enable the debug UART

[Simon Glass]

Add support for the debug UART on link. This is useful for early debugging.

Signed-off-by: Simon Glass <sjg@chromium.org>
---

 arch/x86/cpu/ivybridge/cpu.c      |  7 +++++++
 configs/chromebook_link_defconfig | 10 +++++++---
 2 files changed, 14 insertions(+), 3 deletions(-)

diff --git a/arch/x86/cpu/ivybridge/cpu.c b/arch/x86/cpu/ivybridge/cpu.c
index cce5923..0e6512c 100644
--- a/arch/x86/cpu/ivybridge/cpu.c
+++ b/arch/x86/cpu/ivybridge/cpu.c
@@ -340,3 +340,10 @@ int print_cpuinfo(void)
 
 	return 0;
 }
+
+void board_debug_uart_init(void)
+{
+	/* This enables the debug UART */
+	pci_x86_write_config(NULL, PCH_LPC_DEV, LPC_EN, COMA_LPC_EN,
+			     PCI_SIZE_16);
+}
diff --git a/configs/chromebook_link_defconfig b/configs/chromebook_link_defconfig
index b3ae925..7408cb1 100644
--- a/configs/chromebook_link_defconfig
+++ b/configs/chromebook_link_defconfig
@@ -15,17 +15,21 @@ CONFIG_CMD_BOOTSTAGE=y
 CONFIG_CMD_TPM=y
 CONFIG_CMD_TPM_TEST=y
 CONFIG_OF_CONTROL=y
-CONFIG_DM_PCI=y
-CONFIG_SPI_FLASH=y
 CONFIG_CMD_CROS_EC=y
 CONFIG_CROS_EC=y
 CONFIG_CROS_EC_LPC=y
+CONFIG_SPI_FLASH=y
+CONFIG_DM_PCI=y
+CONFIG_DM_RTC=y
+CONFIG_DEBUG_UART=y
+CONFIG_DEBUG_UART_BASE=0x3f8
+CONFIG_DEBUG_UART_CLOCK=1843200
+CONFIG_DEBUG_UART_BOARD_INIT=y
 CONFIG_DM_TPM=y
 CONFIG_TPM_TIS_LPC=y
 CONFIG_VIDEO_VESA=y
 CONFIG_FRAMEBUFFER_SET_VESA_MODE=y
 CONFIG_FRAMEBUFFER_VESA_MODE_11A=y
-CONFIG_DM_RTC=y
 CONFIG_USE_PRIVATE_LIBGCC=y
 CONFIG_SYS_VSNPRINTF=y
 CONFIG_TPM=y
-- 
2.6.0.rc0.131.gf624c3d

[U-Boot] [PATCH 01/10] x86: chromebook_link: Expand early malloc() memory

[Bin Meng]

On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
> Now that PCI bridges are probed before relocation we need additional memory.
> Each PCI bridge takes 240 bytes at present since it uses the same uclass as
> the PCI controller. Probably we should split this out so that bridges have
> their own uclass.
>
> Expand the memory on link so that it works correctly.
>
> Signed-off-by: Simon Glass <sjg@chromium.org>
> ---
>
>  configs/chromebook_link_defconfig | 1 +
>  1 file changed, 1 insertion(+)
>
> diff --git a/configs/chromebook_link_defconfig b/configs/chromebook_link_defconfig
> index 9855736..b3ae925 100644
> --- a/configs/chromebook_link_defconfig
> +++ b/configs/chromebook_link_defconfig
> @@ -1,4 +1,5 @@
>  CONFIG_X86=y
> +CONFIG_SYS_MALLOC_F_LEN=0x1800
>  CONFIG_VENDOR_GOOGLE=y
>  CONFIG_DEFAULT_DEVICE_TREE="chromebook_link"
>  CONFIG_TARGET_CHROMEBOOK_LINK=y
> --

Reviewed-by: Bin Meng <bmeng.cn@gmail.com>

[U-Boot] [PATCH 02/10] malloc_simple: Add debug() information

[Bin Meng]

On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
> It's useful to get a a trace of memory allocations in early init. Add a
> debug() call to provide that. It can be enabled by adding '#define DEBUG'
> to the top of the file.
>
> Signed-off-by: Simon Glass <sjg@chromium.org>
> ---
>
>  common/malloc_simple.c | 4 ++++
>  1 file changed, 4 insertions(+)
>
> diff --git a/common/malloc_simple.c b/common/malloc_simple.c
> index 134e059..e357827 100644
> --- a/common/malloc_simple.c
> +++ b/common/malloc_simple.c
> @@ -19,10 +19,13 @@ void *malloc_simple(size_t bytes)
>         void *ptr;
>
>         new_ptr = gd->malloc_ptr + bytes;
> +       debug("%s: size=%zx, ptr=%lx, limit=%lx\n", __func__, bytes, new_ptr,
> +             gd->malloc_limit);
>         if (new_ptr > gd->malloc_limit)
>                 return NULL;
>         ptr = map_sysmem(gd->malloc_base + gd->malloc_ptr, bytes);
>         gd->malloc_ptr = ALIGN(new_ptr, sizeof(new_ptr));
> +
>         return ptr;
>  }
>
> @@ -37,6 +40,7 @@ void *memalign_simple(size_t align, size_t bytes)
>                 return NULL;
>         ptr = map_sysmem(addr, bytes);
>         gd->malloc_ptr = ALIGN(new_ptr, sizeof(new_ptr));
> +
>         return ptr;
>  }
>
> --

Reviewed-by: Bin Meng <bmeng.cn@gmail.com>

[U-Boot] [PATCH 03/10] dm: pci: Tidy up auto-config error handling

[Bin Meng]

On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
> When the auto-configuration process fails for a device (generally due to
> lack of memory) we should return the error correctly so that we don't
> continue to try memory allocations which will fail.
>
> Adjust the code to check for errors and abort if something goes wrong.
>
> Signed-off-by: Simon Glass <sjg@chromium.org>
> ---
>
>  drivers/pci/pci-uclass.c | 15 ++++++++++++---
>  1 file changed, 12 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/pci/pci-uclass.c b/drivers/pci/pci-uclass.c
> index b25298f..437c81a 100644
> --- a/drivers/pci/pci-uclass.c
> +++ b/drivers/pci/pci-uclass.c
> @@ -381,9 +381,13 @@ int pci_auto_config_devices(struct udevice *bus)
>              !ret && dev;
>              ret = device_find_next_child(&dev)) {
>                 unsigned int max_bus;
> +               int ret;
>
>                 debug("%s: device %s\n", __func__, dev->name);
> -               max_bus = pciauto_config_device(hose, pci_get_bdf(dev));
> +               ret = pciauto_config_device(hose, pci_get_bdf(dev));
> +               if (ret < 0)
> +                       return ret;
> +               max_bus = ret;
>                 sub_bus = max(sub_bus, max_bus);
>         }
>         debug("%s: done\n", __func__);
> @@ -757,6 +761,8 @@ static int pci_uclass_post_probe(struct udevice *bus)
>
>  #ifdef CONFIG_PCI_PNP
>         ret = pci_auto_config_devices(bus);
> +       if (ret < 0)
> +               return ret;
>  #endif
>
>  #if defined(CONFIG_X86) && defined(CONFIG_HAVE_FSP)
> @@ -773,11 +779,14 @@ static int pci_uclass_post_probe(struct udevice *bus)
>          * Note we only call this 1) after U-Boot is relocated, and 2)
>          * root bus has finished probing.
>          */
> -       if ((gd->flags & GD_FLG_RELOC) && (bus->seq == 0))
> +       if ((gd->flags & GD_FLG_RELOC) && (bus->seq == 0)) {
>                 ret = fsp_init_phase_pci();
> +               if (ret)
> +                       return ret;
> +       }
>  #endif
>
> -       return ret < 0 ? ret : 0;
> +       return 0;
>  }
>
>  static int pci_uclass_child_post_bind(struct udevice *dev)
> --

Reviewed-by: Bin Meng <bmeng.cn@gmail.com>

[U-Boot] [PATCH 04/10] dm: pci: Correct a few debug() statements

[Bin Meng]

On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
> One debug() statement is missing a newline. The other has a repeated word.
> Fix these.
>
> Signed-off-by: Simon Glass <sjg@chromium.org>
> ---
>
>  drivers/pci/pci-uclass.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/pci/pci-uclass.c b/drivers/pci/pci-uclass.c
> index 437c81a..7347160 100644
> --- a/drivers/pci/pci-uclass.c
> +++ b/drivers/pci/pci-uclass.c
> @@ -418,7 +418,7 @@ int dm_pci_hose_probe_bus(struct pci_controller *hose, pci_dev_t bdf)
>
>         ret = device_probe(bus);
>         if (ret) {
> -               debug("%s: Cannot probe bus bus %s: %d\n", __func__, bus->name,
> +               debug("%s: Cannot probe bus %s: %d\n", __func__, bus->name,
>                       ret);
>                 return ret;
>         }
> @@ -537,7 +537,7 @@ static int pci_find_and_bind_driver(struct udevice *parent,
>
>         ret = device_bind_driver(parent, drv, str, devp);
>         if (ret) {
> -               debug("%s: Failed to bind generic driver: %d", __func__, ret);
> +               debug("%s: Failed to bind generic driver: %d\n", __func__, ret);
>                 return ret;
>         }
>         debug("%s: No match found: bound generic driver instead\n", __func__);
> --

Reviewed-by: Bin Meng <bmeng.cn@gmail.com>

[U-Boot] [PATCH 05/10] dm: pci: Adjust pci_find_and_bind_driver() to return -EPERM

[Bin Meng]

On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
> The current code returns 0 even if it failed to find or bind a driver. The
> caller then has to check the returned device to see if it is NULL. It is
> better to return an error code in this case so that it is clear what
> happened.
>
> Adjust the code to return -EPERM, indicating that the device was not bound
> because it is not needed for pre-relocation use. Add comments so that the
> return value is clear.
>
> Signed-off-by: Simon Glass <sjg@chromium.org>
> ---
>
>  drivers/pci/pci-uclass.c | 31 +++++++++++++++++++------------
>  1 file changed, 19 insertions(+), 12 deletions(-)
>
> diff --git a/drivers/pci/pci-uclass.c b/drivers/pci/pci-uclass.c
> index 7347160..54285ee 100644
> --- a/drivers/pci/pci-uclass.c
> +++ b/drivers/pci/pci-uclass.c
> @@ -458,10 +458,17 @@ static bool pci_match_one_id(const struct pci_device_id *id,
>   * pci_find_and_bind_driver() - Find and bind the right PCI driver
>   *
>   * This only looks at certain fields in the descriptor.
> + *
> + * @parent:    Parent bus
> + * @find_id:   Specification of the driver to find
> + * @bdf:       Bus/device/function addreess - see PCI_BDF()
> + * @devp:      Returns a pointer to the device created
> + * @return 0 if OK, -EPERM if the device is not needed before relocation and
> + *        therefore was not created, other -ve value on error
>   */
>  static int pci_find_and_bind_driver(struct udevice *parent,
> -                                   struct pci_device_id *find_id, pci_dev_t bdf,
> -                                   struct udevice **devp)
> +                                   struct pci_device_id *find_id,
> +                                   pci_dev_t bdf, struct udevice **devp)
>  {
>         struct pci_driver_entry *start, *entry;
>         const char *drv;
> @@ -497,7 +504,7 @@ static int pci_find_and_bind_driver(struct udevice *parent,
>                          */
>                         if (!(gd->flags & GD_FLG_RELOC) &&
>                             !(drv->flags & DM_FLAG_PRE_RELOC))
> -                               return 0;
> +                               return -EPERM;
>
>                         /*
>                          * We could pass the descriptor to the driver as
> @@ -525,7 +532,7 @@ static int pci_find_and_bind_driver(struct udevice *parent,
>          * limited (ie: using Cache As RAM).
>          */
>         if (!(gd->flags & GD_FLG_RELOC) && !bridge)
> -               return 0;
> +               return -EPERM;
>
>         /* Bind a generic driver so that the device can be used */
>         sprintf(name, "pci_%x:%x.%x", parent->seq, PCI_DEV(bdf),
> @@ -613,17 +620,17 @@ int pci_bind_bus_devices(struct udevice *bus)
>                         ret = pci_find_and_bind_driver(bus, &find_id, bdf,
>                                                        &dev);
>                 }
> -               if (ret)
> +               if (ret == -EPERM)
> +                       continue;
> +               else if (ret)
>                         return ret;
>
>                 /* Update the platform data */
> -               if (dev) {
> -                       pplat = dev_get_parent_platdata(dev);
> -                       pplat->devfn = PCI_MASK_BUS(bdf);
> -                       pplat->vendor = vendor;
> -                       pplat->device = device;
> -                       pplat->class = class;
> -               }
> +               pplat = dev_get_parent_platdata(dev);
> +               pplat->devfn = PCI_MASK_BUS(bdf);
> +               pplat->vendor = vendor;
> +               pplat->device = device;
> +               pplat->class = class;
>         }
>
>         return 0;
> --

Reviewed-by: Bin Meng <bmeng.cn@gmail.com>

[U-Boot] [PATCH 06/10] debug_uart: Adjust the declaration of debug_uart_init()

[Bin Meng]

Hi Simon,

On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
> We want to be able to add other common code to this function. So change the
> driver's version to have an underscore before it, just like
> _debug_uart_putc(). Define debug_uart_init() to call this version.
>
> Update all drivers to this new method.
>
> Signed-off-by: Simon Glass <sjg@chromium.org>
> ---
>
>  drivers/serial/ns16550.c    | 2 +-
>  drivers/serial/serial_efi.c | 2 +-
>  drivers/serial/serial_s5p.c | 2 +-
>  include/debug_uart.h        | 9 +++++++--
>  lib/efi/efi_stub.c          | 2 +-
>  5 files changed, 11 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/serial/ns16550.c b/drivers/serial/ns16550.c
> index 6275a11..6433844 100644
> --- a/drivers/serial/ns16550.c
> +++ b/drivers/serial/ns16550.c
> @@ -257,7 +257,7 @@ int NS16550_tstc(NS16550_t com_port)
>                         (1 << CONFIG_DEBUG_UART_SHIFT), \
>                 CONFIG_DEBUG_UART_SHIFT)
>
> -void debug_uart_init(void)
> +static inline void _debug_uart_init(void)

Is this 'static inline' change needed? As I don't see the same changes
applied to serial_efi.c and serial_s5p.c

>  {
>         struct NS16550 *com_port = (struct NS16550 *)CONFIG_DEBUG_UART_BASE;
>         int baud_divisor;
> diff --git a/drivers/serial/serial_efi.c b/drivers/serial/serial_efi.c
> index cf57d89..ee3029b 100644
> --- a/drivers/serial/serial_efi.c
> +++ b/drivers/serial/serial_efi.c
> @@ -107,7 +107,7 @@ static int serial_efi_pending(struct udevice *dev, bool input)
>   * There is nothing to init here since the EFI console is already running by
>   * the time we enter U-Boot.
>   */
> -void debug_uart_init(void)
> +void _debug_uart_init(void)
>  {
>  }
>
> diff --git a/drivers/serial/serial_s5p.c b/drivers/serial/serial_s5p.c
> index 3f0b588..e7adf0d 100644
> --- a/drivers/serial/serial_s5p.c
> +++ b/drivers/serial/serial_s5p.c
> @@ -207,7 +207,7 @@ U_BOOT_DRIVER(serial_s5p) = {
>
>  #include <debug_uart.h>
>
> -void debug_uart_init(void)
> +void _debug_uart_init(void)
>  {
>         struct s5p_uart *uart = (struct s5p_uart *)CONFIG_DEBUG_UART_BASE;
>
> diff --git a/include/debug_uart.h b/include/debug_uart.h
> index a75e377..257ba00 100644
> --- a/include/debug_uart.h
> +++ b/include/debug_uart.h
> @@ -38,7 +38,7 @@
>   * To enable the debug UART in your serial driver:
>   *
>   * - #include <debug_uart.h>
> - * - Define debug_uart_init(), trying to avoid using the stack
> + * - Define _debug_uart_init(), trying to avoid using the stack
>   * - Define _debug_uart_putc() as static inline (avoiding stack usage)
>   * - Immediately afterwards, add DEBUG_UART_FUNCS to define the rest of the
>   *     functionality (printch(), etc.)
> @@ -132,6 +132,11 @@ void printhex8(uint value);
>         void printhex8(uint value) \
>         { \
>                 printhex(value, 8); \
> -       }
> +       } \
> +\
> +       void debug_uart_init(void) \
> +       { \
> +               _debug_uart_init(); \
> +       } \
>
>  #endif
> diff --git a/lib/efi/efi_stub.c b/lib/efi/efi_stub.c
> index d4d3e49..e138709 100644
> --- a/lib/efi/efi_stub.c
> +++ b/lib/efi/efi_stub.c
> @@ -59,7 +59,7 @@ struct __packed desctab_info {
>   * considering if we start needing more U-Boot functionality. Note that we
>   * could then move get_codeseg32() to arch/x86/cpu/cpu.c.
>   */
> -void debug_uart_init(void)
> +void _debug_uart_init(void)
>  {
>  }
>
> --

Regards,
Bin

[U-Boot] [PATCH 07/10] debug_uart: Support board-specific UART initialisation

[Bin Meng]

On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
> Some boards need to set things up before the debug UART can be used. On
> these boards a call to debug_uart_init() is insufficient. When this option
> is enabled, the function board_debug_uart_init() will be called when
> debug_uart_init() is called. You can put any code here that is needed to
> set up the UART ready for use, such as set pin multiplexing or enable
> clocks.
>
> Signed-off-by: Simon Glass <sjg@chromium.org>
> ---
>
>  drivers/serial/Kconfig | 11 +++++++++++
>  include/debug_uart.h   | 14 ++++++++++++++
>  2 files changed, 25 insertions(+)
>
> diff --git a/drivers/serial/Kconfig b/drivers/serial/Kconfig
> index ccb80d2..4418a57 100644
> --- a/drivers/serial/Kconfig
> +++ b/drivers/serial/Kconfig
> @@ -109,6 +109,17 @@ config DEBUG_UART_SHIFT
>           value. Use this value to specify the shift to use, where 0=byte
>           registers, 2=32-bit word registers, etc.
>
> +config DEBUG_UART_BOARD_INIT
> +       bool "Enable board-specific debug UART init"
> +       depends on DEBUG_UART
> +       help
> +         Some boards need to set things up before the debug UART can be used.
> +         On these boards a call to debug_uart_init() is insufficient. When
> +         this option is enabled, the function board_debug_uart_init() will
> +         be called when debug_uart_init() is called. You can put any code
> +         here that is needed to set up the UART ready for use, such as set
> +         pin multiplexing or enable clocks.
> +
>  config ROCKCHIP_SERIAL
>         bool "Rockchip on-chip UART support"
>         depends on ARCH_UNIPHIER && DM_SERIAL
> diff --git a/include/debug_uart.h b/include/debug_uart.h
> index 257ba00..a6b7ce8 100644
> --- a/include/debug_uart.h
> +++ b/include/debug_uart.h
> @@ -42,6 +42,11 @@
>   * - Define _debug_uart_putc() as static inline (avoiding stack usage)
>   * - Immediately afterwards, add DEBUG_UART_FUNCS to define the rest of the
>   *     functionality (printch(), etc.)
> + *
> + * If your board needs additional init for the UART to work, enable
> + * CONFIG_DEBUG_UART_BOARD_INIT and write a function called
> + * board_debug_uart_init() to perform that init. When debug_uart_init() is
> + * called, the init will happen automatically.
>   */
>
>  /**
> @@ -57,6 +62,14 @@
>   */
>  void debug_uart_init(void);
>
> +#ifdef CONFIG_DEBUG_UART_BOARD_INIT
> +void board_debug_uart_init(void);
> +#else
> +static inline void board_debug_uart_init(void)
> +{
> +}
> +#endif
> +
>  /**
>   * printch() - Output a character to the debug UART
>   *
> @@ -136,6 +149,7 @@ void printhex8(uint value);
>  \
>         void debug_uart_init(void) \
>         { \
> +               board_debug_uart_init(); \
>                 _debug_uart_init(); \
>         } \
>
> --

Reviewed-by: Bin Meng <bmeng.cn@gmail.com>

[U-Boot] [PATCH 08/10] debug_uart: Add an option to announce the debug UART

[Bin Meng]

On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
> It is useful to see a message from the debug UART early during boot so that
> you know things are working. Add an option to enable this. The message will
> be displayed as soon as debug_uart_init() is called.
>
> Signed-off-by: Simon Glass <sjg@chromium.org>
> ---
>
>  drivers/serial/Kconfig | 10 ++++++++++
>  include/debug_uart.h   |  7 +++++++
>  2 files changed, 17 insertions(+)
>
> diff --git a/drivers/serial/Kconfig b/drivers/serial/Kconfig
> index 4418a57..fa03d17 100644
> --- a/drivers/serial/Kconfig
> +++ b/drivers/serial/Kconfig
> @@ -120,6 +120,16 @@ config DEBUG_UART_BOARD_INIT
>           here that is needed to set up the UART ready for use, such as set
>           pin multiplexing or enable clocks.
>
> +config DEBUG_UART_ANNOUNCE
> +       bool "Show a message when the debug UART starts up"
> +       depends on DEBUG_UART
> +       help
> +         Enable this option to show a message when the debug UART is ready
> +         for use. You will see a message like "<debug_uart> " as soon as
> +         U-Boot has the UART ready for use (i.e. your code calls
> +         debug_uart_init()). This can be useful just as a check that
> +         everything is working.
> +
>  config ROCKCHIP_SERIAL
>         bool "Rockchip on-chip UART support"
>         depends on ARCH_UNIPHIER && DM_SERIAL
> diff --git a/include/debug_uart.h b/include/debug_uart.h
> index a6b7ce8..5d5349b 100644
> --- a/include/debug_uart.h
> +++ b/include/debug_uart.h
> @@ -105,6 +105,12 @@ void printhex4(uint value);
>   */
>  void printhex8(uint value);
>
> +#ifdef CONFIG_DEBUG_UART_ANNOUNCE
> +#define _DEBUG_UART_ANNOUNCE   printascii("<debug_uart> ");
> +#else
> +#define _DEBUG_UART_ANNOUNCE
> +#endif
> +
>  /*
>   * Now define some functions - this should be inserted into the serial driver
>   */
> @@ -151,6 +157,7 @@ void printhex8(uint value);
>         { \
>                 board_debug_uart_init(); \
>                 _debug_uart_init(); \
> +               _DEBUG_UART_ANNOUNCE \
>         } \
>
>  #endif
> --

Reviewed-by: Bin Meng <bmeng.cn@gmail.com>

[U-Boot] [PATCH 09/10] x86: Init the debug UART if enabled

[Bin Meng]

Hi Simon,

On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
> If the debug UART is enabled, get it ready for use at the earliest possible
> opportunity. This is not actually very early, but until we have a stack it
> is difficult to make it work.
>
> Signed-off-by: Simon Glass <sjg@chromium.org>
> ---
>
>  arch/x86/cpu/start.S | 3 +++
>  1 file changed, 3 insertions(+)
>
> diff --git a/arch/x86/cpu/start.S b/arch/x86/cpu/start.S
> index d072825..e6e9cb5 100644
> --- a/arch/x86/cpu/start.S
> +++ b/arch/x86/cpu/start.S
> @@ -134,6 +134,9 @@ car_init_ret:
>          * mov   $'a', %eax
>          * call  printch
>          */

Can we remove the comment block above since it is just duplicate to
the codes below?

> +#ifdef CONFIG_DEBUG_UART
> +       call  debug_uart_init

nits: please use a 'tab' after 'call'.

> +#endif
>
>         /* Get address of global_data */
>         mov     %fs:0, %edx
> --

Regards,
Bin

[U-Boot] [PATCH 10/10] x86: chromebook_link: Enable the debug UART

[Bin Meng]

On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
> Add support for the debug UART on link. This is useful for early debugging.
>
> Signed-off-by: Simon Glass <sjg@chromium.org>
> ---
>
>  arch/x86/cpu/ivybridge/cpu.c      |  7 +++++++
>  configs/chromebook_link_defconfig | 10 +++++++---
>  2 files changed, 14 insertions(+), 3 deletions(-)
>
> diff --git a/arch/x86/cpu/ivybridge/cpu.c b/arch/x86/cpu/ivybridge/cpu.c
> index cce5923..0e6512c 100644
> --- a/arch/x86/cpu/ivybridge/cpu.c
> +++ b/arch/x86/cpu/ivybridge/cpu.c
> @@ -340,3 +340,10 @@ int print_cpuinfo(void)
>
>         return 0;
>  }
> +
> +void board_debug_uart_init(void)
> +{
> +       /* This enables the debug UART */
> +       pci_x86_write_config(NULL, PCH_LPC_DEV, LPC_EN, COMA_LPC_EN,
> +                            PCI_SIZE_16);
> +}
> diff --git a/configs/chromebook_link_defconfig b/configs/chromebook_link_defconfig
> index b3ae925..7408cb1 100644
> --- a/configs/chromebook_link_defconfig
> +++ b/configs/chromebook_link_defconfig
> @@ -15,17 +15,21 @@ CONFIG_CMD_BOOTSTAGE=y
>  CONFIG_CMD_TPM=y
>  CONFIG_CMD_TPM_TEST=y
>  CONFIG_OF_CONTROL=y
> -CONFIG_DM_PCI=y
> -CONFIG_SPI_FLASH=y
>  CONFIG_CMD_CROS_EC=y
>  CONFIG_CROS_EC=y
>  CONFIG_CROS_EC_LPC=y
> +CONFIG_SPI_FLASH=y
> +CONFIG_DM_PCI=y
> +CONFIG_DM_RTC=y
> +CONFIG_DEBUG_UART=y
> +CONFIG_DEBUG_UART_BASE=0x3f8
> +CONFIG_DEBUG_UART_CLOCK=1843200
> +CONFIG_DEBUG_UART_BOARD_INIT=y
>  CONFIG_DM_TPM=y
>  CONFIG_TPM_TIS_LPC=y
>  CONFIG_VIDEO_VESA=y
>  CONFIG_FRAMEBUFFER_SET_VESA_MODE=y
>  CONFIG_FRAMEBUFFER_VESA_MODE_11A=y
> -CONFIG_DM_RTC=y
>  CONFIG_USE_PRIVATE_LIBGCC=y
>  CONFIG_SYS_VSNPRINTF=y
>  CONFIG_TPM=y
> --

Reviewed-by: Bin Meng <bmeng.cn@gmail.com>

[U-Boot] [PATCH 01/10] x86: chromebook_link: Expand early malloc() memory

[Simon Glass]

On 14 September 2015 at 06:15, Bin Meng <bmeng.cn@gmail.com> wrote:
>
> On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
> > Now that PCI bridges are probed before relocation we need additional memory.
> > Each PCI bridge takes 240 bytes at present since it uses the same uclass as
> > the PCI controller. Probably we should split this out so that bridges have
> > their own uclass.
> >
> > Expand the memory on link so that it works correctly.
> >
> > Signed-off-by: Simon Glass <sjg@chromium.org>
> > ---
> >
> >  configs/chromebook_link_defconfig | 1 +
> >  1 file changed, 1 insertion(+)
> >
> > diff --git a/configs/chromebook_link_defconfig b/configs/chromebook_link_defconfig
> > index 9855736..b3ae925 100644
> > --- a/configs/chromebook_link_defconfig
> > +++ b/configs/chromebook_link_defconfig
> > @@ -1,4 +1,5 @@
> >  CONFIG_X86=y
> > +CONFIG_SYS_MALLOC_F_LEN=0x1800
> >  CONFIG_VENDOR_GOOGLE=y
> >  CONFIG_DEFAULT_DEVICE_TREE="chromebook_link"
> >  CONFIG_TARGET_CHROMEBOOK_LINK=y
> > --
>
> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>

Applied to u-boot-x86.

[U-Boot] [PATCH 06/10] debug_uart: Adjust the declaration of debug_uart_init()

[Simon Glass]

Hi Bin,

On 14 September 2015 at 06:16, Bin Meng <bmeng.cn@gmail.com> wrote:
> Hi Simon,
>
> On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
>> We want to be able to add other common code to this function. So change the
>> driver's version to have an underscore before it, just like
>> _debug_uart_putc(). Define debug_uart_init() to call this version.
>>
>> Update all drivers to this new method.
>>
>> Signed-off-by: Simon Glass <sjg@chromium.org>
>> ---
>>
>>  drivers/serial/ns16550.c    | 2 +-
>>  drivers/serial/serial_efi.c | 2 +-
>>  drivers/serial/serial_s5p.c | 2 +-
>>  include/debug_uart.h        | 9 +++++++--
>>  lib/efi/efi_stub.c          | 2 +-
>>  5 files changed, 11 insertions(+), 6 deletions(-)
>>
>> diff --git a/drivers/serial/ns16550.c b/drivers/serial/ns16550.c
>> index 6275a11..6433844 100644
>> --- a/drivers/serial/ns16550.c
>> +++ b/drivers/serial/ns16550.c
>> @@ -257,7 +257,7 @@ int NS16550_tstc(NS16550_t com_port)
>>                         (1 << CONFIG_DEBUG_UART_SHIFT), \
>>                 CONFIG_DEBUG_UART_SHIFT)
>>
>> -void debug_uart_init(void)
>> +static inline void _debug_uart_init(void)
>
> Is this 'static inline' change needed? As I don't see the same changes
> applied to serial_efi.c and serial_s5p.c

Yes I've added it to those two - it is not essential but I think it
makes sense since there is only one declaration and we want to avoid
unnecessary function calls.

>
>>  {
>>         struct NS16550 *com_port = (struct NS16550 *)CONFIG_DEBUG_UART_BASE;
>>         int baud_divisor;
>> diff --git a/drivers/serial/serial_efi.c b/drivers/serial/serial_efi.c
>> index cf57d89..ee3029b 100644
>> --- a/drivers/serial/serial_efi.c
>> +++ b/drivers/serial/serial_efi.c
>> @@ -107,7 +107,7 @@ static int serial_efi_pending(struct udevice *dev, bool input)
>>   * There is nothing to init here since the EFI console is already running by
>>   * the time we enter U-Boot.
>>   */
>> -void debug_uart_init(void)
>> +void _debug_uart_init(void)
>>  {
>>  }
>>
>> diff --git a/drivers/serial/serial_s5p.c b/drivers/serial/serial_s5p.c
>> index 3f0b588..e7adf0d 100644
>> --- a/drivers/serial/serial_s5p.c
>> +++ b/drivers/serial/serial_s5p.c
>> @@ -207,7 +207,7 @@ U_BOOT_DRIVER(serial_s5p) = {
>>
>>  #include <debug_uart.h>
>>
>> -void debug_uart_init(void)
>> +void _debug_uart_init(void)
>>  {
>>         struct s5p_uart *uart = (struct s5p_uart *)CONFIG_DEBUG_UART_BASE;
>>
>> diff --git a/include/debug_uart.h b/include/debug_uart.h
>> index a75e377..257ba00 100644
>> --- a/include/debug_uart.h
>> +++ b/include/debug_uart.h
>> @@ -38,7 +38,7 @@
>>   * To enable the debug UART in your serial driver:
>>   *
>>   * - #include <debug_uart.h>
>> - * - Define debug_uart_init(), trying to avoid using the stack
>> + * - Define _debug_uart_init(), trying to avoid using the stack
>>   * - Define _debug_uart_putc() as static inline (avoiding stack usage)
>>   * - Immediately afterwards, add DEBUG_UART_FUNCS to define the rest of the
>>   *     functionality (printch(), etc.)
>> @@ -132,6 +132,11 @@ void printhex8(uint value);
>>         void printhex8(uint value) \
>>         { \
>>                 printhex(value, 8); \
>> -       }
>> +       } \
>> +\
>> +       void debug_uart_init(void) \
>> +       { \
>> +               _debug_uart_init(); \
>> +       } \
>>
>>  #endif
>> diff --git a/lib/efi/efi_stub.c b/lib/efi/efi_stub.c
>> index d4d3e49..e138709 100644
>> --- a/lib/efi/efi_stub.c
>> +++ b/lib/efi/efi_stub.c
>> @@ -59,7 +59,7 @@ struct __packed desctab_info {
>>   * considering if we start needing more U-Boot functionality. Note that we
>>   * could then move get_codeseg32() to arch/x86/cpu/cpu.c.
>>   */
>> -void debug_uart_init(void)
>> +void _debug_uart_init(void)
>>  {
>>  }
>>
>> --
>
> Regards,
> Bin

Regards,
Simon

[U-Boot] [PATCH 02/10] malloc_simple: Add debug() information

[Simon Glass]

On 14 September 2015 at 06:15, Bin Meng <bmeng.cn@gmail.com> wrote:
> On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
>> It's useful to get a a trace of memory allocations in early init. Add a
>> debug() call to provide that. It can be enabled by adding '#define DEBUG'
>> to the top of the file.
>>
>> Signed-off-by: Simon Glass <sjg@chromium.org>
>> ---
>>
>>  common/malloc_simple.c | 4 ++++
>>  1 file changed, 4 insertions(+)
>>
>> diff --git a/common/malloc_simple.c b/common/malloc_simple.c
>> index 134e059..e357827 100644
>> --- a/common/malloc_simple.c
>> +++ b/common/malloc_simple.c
>> @@ -19,10 +19,13 @@ void *malloc_simple(size_t bytes)
>>         void *ptr;
>>
>>         new_ptr = gd->malloc_ptr + bytes;
>> +       debug("%s: size=%zx, ptr=%lx, limit=%lx\n", __func__, bytes, new_ptr,
>> +             gd->malloc_limit);
>>         if (new_ptr > gd->malloc_limit)
>>                 return NULL;
>>         ptr = map_sysmem(gd->malloc_base + gd->malloc_ptr, bytes);
>>         gd->malloc_ptr = ALIGN(new_ptr, sizeof(new_ptr));
>> +
>>         return ptr;
>>  }
>>
>> @@ -37,6 +40,7 @@ void *memalign_simple(size_t align, size_t bytes)
>>                 return NULL;
>>         ptr = map_sysmem(addr, bytes);
>>         gd->malloc_ptr = ALIGN(new_ptr, sizeof(new_ptr));
>> +
>>         return ptr;
>>  }
>>
>> --
>
> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>

Applied to u-boot-x86.

[U-Boot] [PATCH 03/10] dm: pci: Tidy up auto-config error handling

[Simon Glass]

On 14 September 2015 at 06:15, Bin Meng <bmeng.cn@gmail.com> wrote:
> On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
>> When the auto-configuration process fails for a device (generally due to
>> lack of memory) we should return the error correctly so that we don't
>> continue to try memory allocations which will fail.
>>
>> Adjust the code to check for errors and abort if something goes wrong.
>>
>> Signed-off-by: Simon Glass <sjg@chromium.org>
>> ---
>>
>>  drivers/pci/pci-uclass.c | 15 ++++++++++++---
>>  1 file changed, 12 insertions(+), 3 deletions(-)
[snip]
>
> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>

Applied to u-boot-x86.

[U-Boot] [PATCH 04/10] dm: pci: Correct a few debug() statements

[Simon Glass]

On 14 September 2015 at 06:15, Bin Meng <bmeng.cn@gmail.com> wrote:
> On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
>> One debug() statement is missing a newline. The other has a repeated word.
>> Fix these.
>>
>> Signed-off-by: Simon Glass <sjg@chromium.org>
>> ---
>>
>>  drivers/pci/pci-uclass.c | 4 ++--
>>  1 file changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/pci/pci-uclass.c b/drivers/pci/pci-uclass.c
>> index 437c81a..7347160 100644
>> --- a/drivers/pci/pci-uclass.c
>> +++ b/drivers/pci/pci-uclass.c
>> @@ -418,7 +418,7 @@ int dm_pci_hose_probe_bus(struct pci_controller *hose, pci_dev_t bdf)
>>
>>         ret = device_probe(bus);
>>         if (ret) {
>> -               debug("%s: Cannot probe bus bus %s: %d\n", __func__, bus->name,
>> +               debug("%s: Cannot probe bus %s: %d\n", __func__, bus->name,
>>                       ret);
>>                 return ret;
>>         }
>> @@ -537,7 +537,7 @@ static int pci_find_and_bind_driver(struct udevice *parent,
>>
>>         ret = device_bind_driver(parent, drv, str, devp);
>>         if (ret) {
>> -               debug("%s: Failed to bind generic driver: %d", __func__, ret);
>> +               debug("%s: Failed to bind generic driver: %d\n", __func__, ret);
>>                 return ret;
>>         }
>>         debug("%s: No match found: bound generic driver instead\n", __func__);
>> --
>
> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>

Applied to u-boot-x86.

[U-Boot] [PATCH 05/10] dm: pci: Adjust pci_find_and_bind_driver() to return -EPERM

[Simon Glass]

On 14 September 2015 at 06:15, Bin Meng <bmeng.cn@gmail.com> wrote:
> On Wed, Sep 9, 2015 at 7:52 AM, Simon Glass <sjg@chromium.org> wrote:
>> The current code returns 0 even if it failed to find or bind a driver. The
>> caller then has to check the returned device to see if it is NULL. It is
>> better to return an error code in this case so that it is clear what
>> happened.
>>
>> Adjust the code to return -EPERM, indicating that the device was not bound
>> because it is not needed for pre-relocation use. Add comments so that the
>> return value is clear.
>>
>> Signed-off-by: Simon Glass <sjg@chromium.org>
>> ---
>>
>>  drivers/pci/pci-uclass.c | 31 +++++++++++++++++++------------
>>  1 file changed, 19 insertions(+), 12 deletions(-)
[snip]
> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>

Applied to u-boot-x86.