Kirkwood: add lschlv2 and lsxhl board support
Subsystems (2)
v5 cover Kirkwood: add lschlv2 and lsxhl board support
Changes:
v5:
- combine patchset again. the "net: *" patches should be individually acked by net custodian
- make features configurable at compile time (CONFIG_RANDOM_MACADDR and CONFIG_SETENV_ENETADDR_BY_INDEX)
- remove unused variable in boards/buffalo/lsxl.c
- fix potential compiler warning "too many arguments for format"
- new patch which fixes eth_getenv_enetaddr_by_index() and eth_mac_skip()
- change initial seed of rand() to 1
- enable CONFIG_API and CONFIG_CMD_ELF for lsxl boards
v4:
- typo fixes (thanks Mike)
- seed all 46bits of the generated ethernet address (suggested by Mike)
- split patchset (generic net helpers and lsxl support)
- fix typo in bootcmd_hdd
- removed board/buffalo/lsxl/config.mk patch from patchset
v3:
- add "Kirkwood:" prefix to patch subject
- moved board/buffalo/lsxl/config.mk to an own patch, so it can be separately acked/naked ;)
- removed any hardcoding, that is the mac address is now automatically generated (random, locally administered) and the IP settings are fetched with DHCP/BOOTP.
- add detailed comments to every configuration line in kwbimage.cfg
- add comments in MPP configuration about GPIO usage
- removed lschlv2 ramboot
- use short board ident string
- small cleanups
v2:
- add to buffalo vendor directory instead of Marvell
- add both boards to MAINTAINERS
- don't define values for feature macros
- use tab for vertical alignment
- remove static network configuration, instead introduce a rescue mode
- add some convenience scripts
- small cleanups
Michael Walle (5):
lib: add rand() function
net: add helper to generate random mac address
net: fix potential compiler warning
net: add eth_setenv_enetaddr_by_index()
Kirkwood: add lschlv2 and lsxhl board support
MAINTAINERS | 5 + board/buffalo/lsxl/Makefile | 50 ++++++ board/buffalo/lsxl/kwbimage-lschl.cfg | 229 +++++++++++++++++++++++++ board/buffalo/lsxl/kwbimage-lsxhl.cfg | 229 +++++++++++++++++++++++++ board/buffalo/lsxl/lsxl.c | 302 +++++++++++++++++++++++++++++++++ board/buffalo/lsxl/lsxl.h | 75 ++++++++ boards.cfg | 2 + include/common.h | 6 + include/configs/lsxl.h | 182 ++++++++++++++++++++ include/net.h | 33 ++++ lib/Makefile | 1 + lib/rand.c | 43 +++++ net/eth.c | 51 ++++++- 13 files changed, 1206 insertions(+), 2 deletions(-) create mode 100644 board/buffalo/lsxl/Makefile create mode 100644 board/buffalo/lsxl/kwbimage-lschl.cfg create mode 100644 board/buffalo/lsxl/kwbimage-lsxhl.cfg create mode 100644 board/buffalo/lsxl/lsxl.c create mode 100644 board/buffalo/lsxl/lsxl.h create mode 100644 include/configs/lsxl.h create mode 100644 lib/rand.c
Signature · 1 line
1.7.2.5
PATCH v5 1/5 lib: add rand() function
It's a PRNG using the simple and fast xorshift method.
--- include/common.h | 6 ++++++ lib/Makefile | 1 + lib/rand.c | 43 +++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 50 insertions(+), 0 deletions(-) create mode 100644 lib/rand.c
include/common.h +6 -0
@@ -733,6 +733,12 @@ char * strmhz(char *buf, unsigned long hz);/* lib/crc32.c */ #include <u-boot/crc.h> +/* lib/rand.c */ +#ifdef CONFIG_RANDOM_MACADDR +void srand (unsigned int seed); +unsigned int rand (void ); +#endif +/* common/console.c */ int console_init_f (void );/* Before relocation; uses the serial stuff */ int console_init_r (void );/* After relocation; uses the console stuff */
lib/Makefile +1 -0
@@ -65,6 +65,7 @@ COBJS-y += string.o COBJS-y += time.o COBJS-$(CONFIG_BOOTP_PXE) += uuid.o COBJS-y += vsprintf.o +COBJS-$(CONFIG_RANDOM_MACADDR) += rand.o COBJS := $(COBJS-y) SRCS := $(COBJS:.o=.c)
lib/rand.c new file +43 -0
43 lines not on this page · raw message
Signature · 1 line
1.7.2.5
PATCH v5 1/5 lib: add rand() function
Am Samstag 12 Mai 2012, 00:50:45 schrieb Michael Walle:
Quoted 4 lines
It's a PRNG using the simple and fast xorshift method.
Signed-off-by Michael Walle <[email protected]>Cc Wolfgang Denk <[email protected]>
Hi Wolfgang,
can you ack this patch? In that case i could add the acked-by line in the next version of this patch series to ease the picking by Prafulla. All your previous comments should be addressed :)
Quoted 13 lines
--- include/common.h | 6 ++++++ lib/Makefile | 1 + lib/rand.c | 43 +++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 50 insertions(+), 0 deletions(-) create mode 100644 lib/rand.cinclude/common.h +0
-0@@ -733,6 +733,12 @@ char * strmhz(char *buf, unsigned long hz);/* lib/crc32.c */
include/common.h char * strmhz(char *buf, unsigned long hz); · line 743
Quoted 10 lines
include/common.h +6
-0#include <u-boot/crc.h> +/* lib/rand.c */ +#ifdef CONFIG_RANDOM_MACADDR +void srand (unsigned int seed); +unsigned int rand (void ); +#endif +/* common/console.c */ int console_init_f (void );/* Before relocation; uses the serial stuff
*/
include/common.h char * strmhz(char *buf, unsigned long hz); · line 744
Quoted 1 line
include/common.h +0
-0int console_init_r (void );/* After relocation; uses the console stuff
*/
Quoted 61 lines
lib/Makefile +1
-0@@ -65,6 +65,7 @@ COBJS-y += string.o COBJS-y += time.o COBJS-$(CONFIG_BOOTP_PXE) += uuid.o COBJS-y += vsprintf.o +COBJS-$(CONFIG_RANDOM_MACADDR) += rand.o COBJS := $(COBJS-y) SRCS := $(COBJS:.o=.c)lib/rand.c new file +43
-0@@ -0,0 +1,43 @@ +/* +* Simple xorshift PRNG +* see http://www.jstatsoft.org/v08/i14/paper +* +* Copyright (c) 2012 Michael Walle +* Michael Walle <[email protected]> +* +* See file CREDITS for list of people who contributed to this +* project. +* +* This program is free software; you can redistribute it and/or +* modify it under the terms of the GNU General Public License as +* published by the Free Software Foundation; either version 2 of +* the License, or (at your option) any later version. +* +* This program is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +* GNU General Public License for more details. +* +* You should have received a copy of the GNU General Public License +* along with this program; if not, write to the Free Software +* Foundation, Inc., 59 Temple Place, Suite 330, Boston, +* MA 02111-1307 USA +*/ + +#include <common.h> + +static unsigned int y =1U ; + +void srand (unsigned int seed) +{ + y = seed; +} + +unsigned int rand (void ) +{ + y ^= (y <<13 ); + y ^= (y >>17 ); + y ^= (y <<5 ); + +return y; +}
Signature · 1 line
Michael
PATCH v5 1/5 lib: add rand() function
Quoted 16 lines
-----Original Message-----
From: Michael Walle [mailto:michael at walle.cc]
Sent: 23 May 2012 22:05
To: u-boot at lists.denx.de
Cc: Prafulla Wadaskar; Wolfgang Denk
Subject: Re: [PATCH v5 1/5] lib: addrand()functionAm Samstag 12 Mai 2012, 00:50:45 schrieb Michael Walle:
Quoted 4 lines
It's a PRNG using the simple and fast xorshift method.
Signed-off-by Michael Walle <[email protected]>Cc Wolfgang Denk <[email protected]>
Quoted 15 lines
Hi Wolfgang,
can you ack this patch? In that case i could add the acked-by line in the next
version of this patch series to ease the picking by Prafulla. All your previous comments should be addressed :)Quoted 7 lines
--- include/common.h | 6 ++++++ lib/Makefile | 1 + lib/rand.c | 43 +++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 50 insertions(+), 0 deletions(-) create mode 100644 lib/rand.c
If I consider the files being checked in, those are generic, I would suggest if Wolfgang can pull them.
Regards..
Prafulla . . .
PATCH v5 1/5 lib: add rand() function
Quoted 16 lines · Prafulla Wadaskar
Quoted 15 lines
Hi Wolfgang,
can you ack this patch? In that case i could add the acked-by line in the next
version of this patch series to ease the picking by Prafulla. All your previous comments should be addressed :)Quoted 7 lines
--- include/common.h | 6 ++++++ lib/Makefile | 1 + lib/rand.c | 43 +++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 50 insertions(+), 0 deletions(-) create mode 100644 lib/rand.c
Quoted 2 lines
If I consider the files being checked in, those are generic, I would suggest if Wolfgang can pull them.
Wolfgang won't pull this alone, see here:
http://lists.denx.de/pipermail/u-boot/2012-May/124191.html
and here
http://lists.denx.de/pipermail/u-boot/2012-May/124202.html
Signature · 1 line
michael
PATCH v5 2/5 net: add helper to generate random mac address
In reply to Michael WalleAdd new function eth_random_enetaddr() to generate a locally administered ethernet address.
--- include/net.h | 17 +++++++++++++++++ net/eth.c | 22 ++++++++++++++++++++++ 2 files changed, 39 insertions(+), 0 deletions(-)
include/net.h +17 -0
@@ -118,6 +118,23 @@ extern int eth_setenv_enetaddr(char *name, const uchar *enetaddr);extern int eth_getenv_enetaddr_by_index (const char *base_name,int index,uchar *enetaddr); +#ifdef CONFIG_RANDOM_MACADDR +/* +* The u-boot policy does not allow hardcoded ethernet addresses. Under the +* following circumstances a random generated address is allowed: +* - in emergency cases, where you need a working network connection to set +* the ethernet address. +* Eg. you want a rescue boot and don't have a serial port to access the +* CLI to set environment variables. +* +* In these cases, we generate a random locally administered ethernet address. +* +* Args: +* enetaddr - returns 6 byte hardware address +*/ +extern void eth_random_enetaddr (uchar *enetaddr); +#endif +extern int usb_eth_initialize (bd_t *bi);extern int eth_init (bd_t *bis);/* Initialize the device */ extern int eth_send (volatile void *packet,int length);/* Send a packet */
net/eth.c +22 -0
@@ -70,6 +70,28 @@ static int eth_mac_skip(int index)return ((skip_state =getenv (enetvar)) !=NULL ); } +#ifdef CONFIG_RANDOM_MACADDR +void eth_random_enetaddr (uchar *enetaddr) +{ +uint32_t rval; + +srand (get_timer (0 )); + + rval =rand (); + enetaddr[0 ] = rval &0xff ; + enetaddr[1 ] = (rval >>8 ) &0xff ; + enetaddr[2 ] = (rval >>16 ) &0xff ; + + rval =rand (); + enetaddr[3 ] = rval &0xff ; + enetaddr[4 ] = (rval >>8 ) &0xff ; + enetaddr[5 ] = (rval >>16 ) &0xff ; + +/* make sure it's local and unicast */ + enetaddr[0 ] = (enetaddr[0 ] |0x02 ) & ~0x01 ; +} +#endif +/* * CPU and board-specific Ethernet initializations. Aliased function * signals caller to move on
Signature · 1 line
1.7.2.5
PATCH v5 2/5 net: add helper to generate random mac address
Quoted 5 lines · Michael Walle
Add new function
eth_random_enetaddr()to generate a locally administered ethernet address.Signed-off-by Michael Walle <[email protected]>Cc Joe Hershberger <[email protected]>
PATCH v5 3/5 net: fix potential compiler warning
In reply to Michael WalleFuture compiler versions may generate a "too many arguments for functions" warning.
--- net/eth.c | 14 ++++++++++++-- 1 files changed, 12 insertions(+), 2 deletions(-)
net/eth.c +12 -2
@@ -58,7 +58,12 @@ int eth_getenv_enetaddr_by_index(const char *base_name, int index, uchar *enetaddr) {char enetvar[32 ]; -sprintf (enetvar, index ?"%s%daddr" :"%saddr" , base_name, index); + +if (index) +sprintf (enetvar,"%s%daddr" , base_name, index); +else +sprintf (enetvar,"%saddr" , base_name); +return eth_getenv_enetaddr (enetvar, enetaddr); }
@@ -66,7 +71,12 @@ static int eth_mac_skip(int index) {char enetvar[15 ];char *skip_state; -sprintf (enetvar, index ?"eth%dmacskip" :"ethmacskip" , index); + +if (index) +sprintf (enetvar,"eth%dmacskip" , index); +else +sprintf (enetvar,"ethmacskip" ); +return ((skip_state =getenv (enetvar)) !=NULL ); }
Signature · 1 line
1.7.2.5
PATCH v5 3/5 net: fix potential compiler warning
Hi Michael,
Quoted 30 lines · Michael Walle
Future compiler versions may generate a "too many arguments for functions" warning.
Signed-off-by Michael Walle <[email protected]>Cc Joe Hershberger <[email protected]>---
?net/eth.c | ? 14 ++++++++++++--
?1 files changed, 12 insertions(+), 2 deletions(-)net/eth.c +6
-1@@ -58,7 +58,12 @@ int eth_getenv_enetaddr_by_index(const char *base_name, int index, ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? uchar *enetaddr) ?{ ? ? ? ?char enetvar[32]; - ? ? ? sprintf(enetvar, index ? "%s%daddr" : "%saddr", base_name, index); + + ? ? ? if (index) + ? ? ? ? ? ? ? sprintf(enetvar, "%s%daddr", base_name, index); + ? ? ? else + ? ? ? ? ? ? ? sprintf(enetvar, "%saddr", base_name); + ? ? ? ?return eth_getenv_enetaddr(enetvar, enetaddr); ?}@@ -66,7 +71,12 @@ static int eth_mac_skip(int index) ?{ ? ? ? ?char enetvar[15];
net/eth.c static int eth_mac_skip(int index) · line 81
Quoted 10 lines
net/eth.c +6
-1? ? ? ?char *skip_state; - ? ? ? sprintf(enetvar, index ? "eth%dmacskip" : "ethmacskip", index); + + ? ? ? if (index) + ? ? ? ? ? ? ? sprintf(enetvar, "eth%dmacskip", index); + ? ? ? else + ? ? ? ? ? ? ? sprintf(enetvar, "ethmacskip"); + ? ? ? ?return ((skip_state = getenv(enetvar)) != NULL); ?}
This seems like this could improve by consolidating the logic to come up with the ethaddr variable name in an inline function. Especially since you are about to add another use in the setenv case. Do you agree, Mike? Or is 3 uses of a relatively simple algorithm better inlined than separated?
-Joe
PATCH v5 4/5 net: add eth_setenv_enetaddr_by_index()
In reply to Michael Walle--- include/net.h | 16 ++++++++++++++++ net/eth.c | 15 +++++++++++++++ 2 files changed, 31 insertions(+), 0 deletions(-)
include/net.h +16 -0
@@ -104,7 +104,9 @@ extern struct eth_device *eth_get_dev_by_index(int index); /* get dev @ index */extern int eth_get_dev_index (void );/* get the device index */ extern void eth_parse_enetaddr (const char *addr,uchar *enetaddr);extern int eth_getenv_enetaddr (char *name,uchar *enetaddr); +#ifdef CONFIG_SETENV_ENETADDR_BY_INDEX extern int eth_setenv_enetaddr (char *name,const uchar *enetaddr); +#endif /* * Get the hardware address for an ethernet interface .
@@ -118,6 +120,20 @@ extern int eth_setenv_enetaddr(char *name, const uchar *enetaddr);extern int eth_getenv_enetaddr_by_index (const char *base_name,int index,uchar *enetaddr); +#ifdef CONFIG_SETENV_ENETADDR_BY_INDEX +/* +* Set the hardware address for an ethernet interface . +* Args: +* base_name - base name for device (normally "eth") +* index - device index number (0 for first) +* enetaddr - returns 6 byte hardware address +* Returns: +* 0 on success, else 1. +*/ +extern int eth_setenv_enetaddr_by_index (const char *base_name,int index, +const uchar *enetaddr); +#endif +#ifdef CONFIG_RANDOM_MACADDR /* * The u-boot policy does not allow hardcoded ethernet addresses. Under the
net/eth.c +15 -0
@@ -67,6 +67,21 @@ int eth_getenv_enetaddr_by_index(const char *base_name, int index,return eth_getenv_enetaddr (enetvar, enetaddr); } +#ifdef CONFIG_SETENV_ENETADDR_BY_INDEX +int eth_setenv_enetaddr_by_index (const char *base_name,int index, +const uchar *enetaddr) +{ +char enetvar[32 ]; + +if (index) +sprintf (enetvar,"%s%daddr" , base_name, index); +else +sprintf (enetvar,"%saddr" , base_name); + +return eth_setenv_enetaddr (enetvar, enetaddr); +} +#endif +static int eth_mac_skip (int index) {char enetvar[15 ];
Signature · 1 line
1.7.2.5
PATCH v5 4/5 net: add eth_setenv_enetaddr_by_index()
Am Samstag 12 Mai 2012, 00:50:48 schrieb Michael Walle:
include/net.h extern struct eth_device *eth_get_dev_by_index(int · line 104
Quoted 13 lines
Signed-off-by Michael Walle <[email protected]>Cc Joe Hershberger <[email protected]>--- include/net.h | 16 ++++++++++++++++ net/eth.c | 15 +++++++++++++++ 2 files changed, 31 insertions(+), 0 deletions(-)include/net.h +0
-0@@ -104,7 +104,9 @@ extern struct eth_device *eth_get_dev_by_index(int index); /* get dev @ index */ extern int eth_get_dev_index (void);
/* get
include/net.h extern struct eth_device *eth_get_dev_by_index(int index); /* get dev @ index */ · line 109
Quoted 6 lines
the device index */
extern voideth_parse_enetaddr(const char *addr, uchar *enetaddr); extern inteth_getenv_enetaddr(char *name, uchar *enetaddr);include/net.h +2
-0+#ifdef CONFIG_SETENV_ENETADDR_BY_INDEX extern int eth_setenv_enetaddr (char *name,const uchar *enetaddr); +#endif
sorry, wrong function, will be corrected in next version.
--
michael
PATCH v5 4/5 net: add eth_setenv_enetaddr_by_index()
In reply to Michael WalleHi Michael,
Quoted 12 lines · Michael Walle
Signed-off-by Michael Walle <[email protected]>Cc Joe Hershberger <[email protected]>---
?include/net.h | ? 16 ++++++++++++++++ ?net/eth.c ? ? | ? 15 +++++++++++++++
?2 files changed, 31 insertions(+), 0 deletions(-)include/net.h +0
-0@@ -104,7 +104,9 @@ extern struct eth_device *eth_get_dev_by_index(int index); /* get dev @ index */
Quoted 6 lines
?extern int
eth_get_dev_index(void); ? ? ? ? ? /* get the device index */ ?extern voideth_parse_enetaddr(const char *addr, uchar *enetaddr); ?extern inteth_getenv_enetaddr(char *name, uchar *enetaddr); +#ifdefCONFIG_SETENV_ENETADDR_BY_INDEX
?extern inteth_setenv_enetaddr(char *name, const uchar *enetaddr); +#endif
Which other function did you intend? You already guarded the only definition whose implementation is guarded.
include/net.h extern int eth_setenv_enetaddr(char *name, const uchar *enetaddr); · line 123
Quoted 7 lines
?/*
?* Get the hardware address for an ethernet interface .include/net.h +1
-0@@ -118,6 +120,20 @@ extern int eth_setenv_enetaddr(char *name, const uchar *enetaddr); ?extern int eth_getenv_enetaddr_by_index(const char *base_name, int index, ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?uchar *enetaddr); +#ifdef CONFIG_SETENV_ENETADDR_BY_INDEX
Personally I don't like config options for something so tiny. If you really must guard against including it, use the CONFIG_RANDOM_MACADDR instead. I'd prefer to just let the linker exclude it when it's never referenced.
Quoted 28 lines
include/net.h +13
-0+/* +* Set the hardware address for an ethernet interface . +* Args: +* ? ? base_name - base name for device (normally "eth") +* ? ? index - device index number (0 for first) +* ? ? enetaddr - returns 6 byte hardware address +* Returns: +* ? ? 0 on success, else 1. +*/ +extern int eth_setenv_enetaddr_by_index (const char *base_name,int index, + ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? const uchar *enetaddr); +#endif +?#ifdef
CONFIG_RANDOM_MACADDR
?/*
?* The u-boot policy does not allow hardcoded ethernet addresses. Under thenet/eth.c +4
-0@@ -67,6 +67,21 @@ int eth_getenv_enetaddr_by_index(const char *base_name, int index, ? ? ? ?return eth_getenv_enetaddr(enetvar, enetaddr); ?} +#ifdef CONFIG_SETENV_ENETADDR_BY_INDEX +int eth_setenv_enetaddr_by_index (const char *base_name,int index, + ? ? ? ? ? ? ? const uchar *enetaddr) +{
net/eth.c int eth_getenv_enetaddr_by_index(const char *base_name, int index, · line 83
Quoted 10 lines
net/eth.c +10
-0+ ? ? ? char enetvar[32]; + + ? ? ? if (index) + ? ? ? ? ? ? ? sprintf(enetvar, "%s%daddr", base_name, index); + ? ? ? else + ? ? ? ? ? ? ? sprintf(enetvar, "%saddr", base_name); + + ? ? ? return eth_setenv_enetaddr(enetvar, enetaddr); +} +#endif
You should synchronize with Rob Herring who is also trying to add this eth_setenv_enetaddr_by_index() function.
http://patchwork.ozlabs.org/patch/152590/
Also since Rob's patch would use this function all the time, the CONFIG_SETENV_ENETADDR_BY_INDEX would be removed, right?
Thanks,
-Joe
PATCH v5 4/5 net: add eth_setenv_enetaddr_by_index()
Hi Joe,
Am Mittwoch 16 Mai 2012, 02:56:39 schrieb Joe Hershberger:
Quoted 10 lines
Hi Michael,
Quoted 7 lines · Michael Walle
Signed-off-by Michael Walle <[email protected]>Cc Joe Hershberger <[email protected]>--- include/net.h | 16 ++++++++++++++++ net/eth.c | 15 +++++++++++++++ 2 files changed, 31 insertions(+), 0 deletions(-)
Quoted 15 lines
include/net.h extern struct eth_device *eth_get_dev_by_index(int · line 110
Quoted 12 lines
include/net.h +2
-0@@ -104,7 +104,9 @@ extern struct eth_device *eth_get_dev_by_index(int index); /* get dev @ index */ extern int eth_get_dev_index (void);/* get the device index */ extern void eth_parse_enetaddr (const char *addr,uchar *enetaddr);extern int eth_getenv_enetaddr (char *name,uchar *enetaddr); +#ifdef CONFIG_SETENV_ENETADDR_BY_INDEX extern int eth_setenv_enetaddr (char *name,const uchar *enetaddr); +#endif Which other function did you intend? You already guarded the only definition whose implementation is guarded.
That was a mistake. Will be fixed in the next version.
Quoted 11 lines
Quoted 7 lines
include/net.h +0
-0/* * Get the hardware address for an ethernet interface . @@ -118,6 +120,20 @@ extern int eth_setenv_enetaddr(char *name, constuchar *enetaddr); extern int
eth_getenv_enetaddr_by_index(const char *base_name, int index, uchar *enetaddr);+#ifdef
CONFIG_SETENV_ENETADDR_BY_INDEXPersonally I don't like config options for something so tiny. If you really must guard against including it, use the
CONFIG_RANDOM_MACADDRinstead.
Me, too. But Wolfgang doesn't accept patches which increases the code size for other boards.
I don't think it is a good idea to use CONFIG_RANDOM_MACADDR for such a generic function, which will (hopefully) be used by others too.
Quoted 2 lines
I'd prefer to just let the linker exclude it when it's never referenced.
Correct, me if i'm wrong, but the linker can only discard unused sections. But there was some discussion on -ffunction-sections and -fdata-sections on the ML. So maybe theres some light at the end of the tunnel ;)
Quoted 35 lines
Quoted 35 lines
include/net.h +10
-0+/* +* Set the hardware address for an ethernet interface . +* Args: +* base_name - base name for device (normally "eth") +* index - device index number (0 for first) +* enetaddr - returns 6 byte hardware address +* Returns: +* 0 on success, else 1. +*/ +extern int eth_setenv_enetaddr_by_index(const char *base_name, intindex, + const uchar *enetaddr);
include/net.h +2
-0+#endif +#ifdef CONFIG_RANDOM_MACADDR /* * The u-boot policy does not allow hardcoded ethernet addresses. Under the diff --git
a/net/eth.cb/net/eth.c
index afce863..d66e22a 100644net/eth.c +10
-0@@ -67,6 +67,21 @@ int eth_getenv_enetaddr_by_index(const char *base_name, int index, return eth_getenv_enetaddr(enetvar, enetaddr); } +#ifdef CONFIG_SETENV_ENETADDR_BY_INDEX +int eth_setenv_enetaddr_by_index (const char *base_name,int index, +const uchar *enetaddr) +{ +char enetvar[32 ]; + +if (index) +sprintf (enetvar,"%s%daddr" , base_name, index); +else +sprintf (enetvar,"%saddr" , base_name);
Quoted 8 lines
net/eth.c int eth_getenv_enetaddr_by_index(const char *base_name, int index, · line 83
Quoted 4 lines
net/eth.c +4
-0+ + return eth_setenv_enetaddr(enetvar, enetaddr); +} +#endif You should synchronize with Rob Herring who is also trying to add this
eth_setenv_enetaddr_by_index()function.
http://patchwork.ozlabs.org/patch/152590/
ah thanks for the hint.
Quoted 3 lines
Also since Rob's patch would use this function all the time, the
CONFIG_SETENV_ENETADDR_BY_INDEXwould be removed, right?
right, but see above.
Signature · 1 line
Michael
PATCH v5 4/5 net: add eth_setenv_enetaddr_by_index()
Hi Michael,
Quoted 18 lines · Michael Walle
Hi Joe,
Am Mittwoch 16 Mai 2012, 02:56:39 schrieb Joe Hershberger:
Quoted 14 lines
Hi Michael,
Quoted 11 lines · Michael Walle
Signed-off-by Michael Walle <[email protected]>Cc Joe Hershberger <[email protected]>----
net/eth.c +0
-0@@ -67,6 +67,21 @@ int eth_getenv_enetaddr_by_index(const char *base_name, int index, return eth_getenv_enetaddr(enetvar, enetaddr); ?}
Quoted 14 lines
Quoted 6 lines
Quoted 6 lines
net/eth.c +6
-0+#ifdef CONFIG_SETENV_ENETADDR_BY_INDEX +int eth_setenv_enetaddr_by_index (const char *base_name,int index, + ? ? ? ? ? ? ? const uchar *enetaddr) +{ + ? ? ? char enetvar[32]; +Quoted 8 lines
net/eth.c int eth_getenv_enetaddr_by_index(const char · line 83
Quoted 8 lines
net/eth.c +8
-0+ ? ? ? if (index) + ? ? ? ? ? ? ? sprintf(enetvar, "%s%daddr", base_name, index); + ? ? ? else + ? ? ? ? ? ? ? sprintf(enetvar, "%saddr", base_name); + + ? ? ? return eth_setenv_enetaddr(enetvar, enetaddr); +} +#endif
How about something like this instead...
-----------8<-----------8<-----------8<-----------8<-----------
include/net.h +16 -0
@@ -600,6 +600,22 @@ static inline int is_broadcast_ether_addr(const u8 *addr) }/* +* Create a variable name for the ethernet address +* +* @param base_name Input string constant (typically "eth") prepended to name +* @param index Index to be added to the name if not 0 +* @param ethaddr_var Output string variable name for MAC address +*/ +static inline void eth_enetaddr_var_by_index (const char *base_name,int index, +char *ethaddr_var) +{ +if (index) +sprintf (ethaddr_var,"%s%daddr" , base_name, index); +else +sprintf (ethaddr_var,"%saddr" , base_name); +} + +/* * is_valid_ether_addr - Determine if the given Ethernet address is valid * @addr: Pointer to a six-byte array containing the Ethernet address *
net/eth.c +0 -0
@@ -58,7 +58,8 @@ int eth_getenv_enetaddr_by_index(const char
*base_name, int index,
uchar *enetaddr)
{
char enetvar[32];
- sprintf(enetvar, index ? "%s%daddr" : "%saddr", base_name, index);
+
+ eth_enetaddr_var_by_index(base_name, index, enetvar);
return eth_getenv_enetaddr(enetvar, enetaddr);
}
----------->8----------->8----------->8----------->8-----------
And then in your board support...
-----------8<-----------8<-----------8<-----------8<-----------
board/buffalo/lsxl/lsxl.c new file +0 -0
+...
+
+static void rescue_mode(void)
+{
+ uchar enetaddr[6];
+ char enetvar[32];
+
+ printf("Entering rescue mode..\n");
+ if (!eth_getenv_enetaddr_by_index("eth", 0, enetaddr)) {
+ eth_random_enetaddr(enetaddr);
+ eth_enetaddr_var_by_index("eth", 0, enetvar);
+ if (eth_setenv_enetaddr(enetvar, enetaddr)) {
+ printf("Failed to set ethernet address\n");
+ set_led(LED_ALARM_BLINKING);
+ return;
+ }
+ }
+ setenv("bootsource", "rescue");
+}
+
----------->8----------->8----------->8----------->8-----------
That way you aren't adding a function that noone else uses and you don't need a silly guard around it.
Cheers,
-Joe
PATCH v5 4/5 net: add eth_setenv_enetaddr_by_index()
Am Freitag 25 Mai 2012, 20:50:09 schrieb Joe Hershberger:
[..snip..]
Quoted 2 lines
That way you aren't adding a function that noone else uses and you don't need a silly guard around it.
Hi Joe,
thanks for the review. I think i'll drop the dynamic "ethernet variable name" support entirely. Two reasons:
- it isn't really dynamic, eg. the function always evaluates the name to "ethaddr"
- back then when i wrote the support for the board the name was "eth1addr" and mike suggested to use the by index function for the getter and a new function for the setter. But i guess its not worth the hassle ;)
btw, imho your solution introduces a discrepancy between the setter and getter. i hope it doesnt bother you if i don't adapt your solution. But thanks for the work.
Signature · 1 line
Michael
PATCH v5 4/5 net: add eth_setenv_enetaddr_by_index()
Hi Michael,
Quoted 7 lines · Michael Walle
Am Freitag 25 Mai 2012, 20:50:09 schrieb Joe Hershberger:
[..snip..]Quoted 2 lines
That way you aren't adding a function that noone else uses and you don't need a silly guard around it.
Hi Joe,
Quoted 7 lines
thanks for the review. I think i'll drop the dynamic "ethernet variable name" support entirely. Two reasons:
?- it isn't really dynamic, eg. the function always evaluates the name to ? "ethaddr"
?- back then when i wrote the support for the board the name was "eth1addr" ? and mike suggested to use the by index function for the getter and a new ? function for the setter. But i guess its not worth the hassle ;)
OK. Since it is specific to your board support anyway, that seems fine.
Quoted 2 lines
btw, imho your solution introduces a discrepancy between the setter and getter.
It does, yes. There is an inherent discrepancy, though, given that one is used and the other is not. I have no problem adding the accessor if there are clients for it to justify its code space.
Quoted 2 lines
i hope it doesnt bother you if i don't adapt your solution. But thanks for the work.
Not a problem at all.
-Joe
PATCH v5 4/5 net: add eth_setenv_enetaddr_by_index()
In reply to Joe HershbergerAm Mittwoch 16 Mai 2012, 02:56:39 schrieb Joe Hershberger:
Quoted 3 lines
You should synchronize with Rob Herring who is also trying to add this
eth_setenv_enetaddr_by_index()function.
http://patchwork.ozlabs.org/patch/152590/
Hi Rob,
Joe mentioned you have the same patch as me here:
http://patchwork.ozlabs.org/patch/158624/
What is the current status of your patch?
Wolfgang had some comments on my patch:
http://patchwork.ozlabs.org/patch/158354/
Could you incorporate these changes into your patch?
Unfortunately, i don't see a way to add the helper function without meeting Wolfgangs requirement not to increase the binary file for other boards. Unless gc-sections and function-sections are used on ARM, or adding an CONFIG_XX macro to mask the helper on compile time.
Signature · 1 line
Michael
PATCH v5 4/5 net: add eth_setenv_enetaddr_by_index()
Quoted 11 lines · Michael Walle
Am Mittwoch 16 Mai 2012, 02:56:39 schrieb Joe Hershberger:
Quoted 3 lines
You should synchronize with Rob Herring who is also trying to add this
eth_setenv_enetaddr_by_index()function.
http://patchwork.ozlabs.org/patch/152590/Hi Rob,
Joe mentioned you have the same patch as me here:
http://patchwork.ozlabs.org/patch/158624/What is the current status of your patch?
I have no idea. It's been in the patch system for a month, gotten no comments, and delegated to jhersh.
Quoted 9 lines
Wolfgang had some comments on my patch:
http://patchwork.ozlabs.org/patch/158354/Could you incorporate these changes into your patch?
Unfortunately, i don't see a way to add the helper function without meeting Wolfgangs requirement not to increase the binary file for other boards. Unless gc-sections and function-sections are used on ARM, or adding an
CONFIG_XXmacro to mask the helper on compile time.
There was a patch series for gc-sections recently...
I don't see a way to meet this impossible requirement either, and I don't really have the bandwidth right now.
Rob
PATCH v5 4/5 net: add eth_setenv_enetaddr_by_index()
Am Dienstag 22 Mai 2012, 23:41:55 schrieb Rob Herring:
Quoted 19 lines · Michael Walle
Quoted 11 lines
Am Mittwoch 16 Mai 2012, 02:56:39 schrieb Joe Hershberger:
Quoted 3 lines
You should synchronize with Rob Herring who is also trying to add this
eth_setenv_enetaddr_by_index()function.
http://patchwork.ozlabs.org/patch/152590/Hi Rob,
Joe mentioned you have the same patch as me here:
http://patchwork.ozlabs.org/patch/158624/What is the current status of your patch?
I have no idea. It's been in the patch system for a month, gotten no comments, and delegated to jhersh.
Quoted 3 lines
Wolfgang had some comments on my patch:
http://patchwork.ozlabs.org/patch/158354/
Quoted 8 lines
Quoted 6 lines
Could you incorporate these changes into your patch?
Unfortunately, i don't see a way to add the helper function without meeting Wolfgangs requirement not to increase the binary file for other boards. Unless gc-sections and function-sections are used on ARM, or adding an
CONFIG_XXmacro to mask the helper on compile time.There was a patch series for gc-sections recently...
yeah saw that too, but someone mentioned there were problems on arm, iirc.
Quoted 2 lines
I don't see a way to meet this impossible requirement either, and I don't really have the bandwidth right now.
ok, thanks for the clarification :)
Signature · 1 line
Michael
PATCH v5 5/5 Kirkwood: add lschlv2 and lsxhl board support
In reply to Michael WalleThis patch adds support for both the Linkstation Live (LS-CHLv2) and Linkstation Pro (LS-XHL) by Buffalo.
--- MAINTAINERS | 5 + board/buffalo/lsxl/Makefile | 50 ++++++ board/buffalo/lsxl/kwbimage-lschl.cfg | 229 +++++++++++++++++++++++++ board/buffalo/lsxl/kwbimage-lsxhl.cfg | 229 +++++++++++++++++++++++++ board/buffalo/lsxl/lsxl.c | 302 +++++++++++++++++++++++++++++++++ board/buffalo/lsxl/lsxl.h | 75 ++++++++ boards.cfg | 2 + include/configs/lsxl.h | 182 ++++++++++++++++++++ 8 files changed, 1074 insertions(+), 0 deletions(-) create mode 100644 board/buffalo/lsxl/Makefile create mode 100644 board/buffalo/lsxl/kwbimage-lschl.cfg create mode 100644 board/buffalo/lsxl/kwbimage-lsxhl.cfg create mode 100644 board/buffalo/lsxl/lsxl.c create mode 100644 board/buffalo/lsxl/lsxl.h create mode 100644 include/configs/lsxl.h
MAINTAINERS +5 -0
@@ -894,6 +894,11 @@ Prafulla Wadaskar <[email protected]> rd6281a ARM926EJS (Kirkwood SoC) sheevaplug ARM926EJS (Kirkwood SoC) +Michael Walle <[email protected]> + + lschlv2 ARM926EJS (Kirkwood SoC) + lsxhl ARM926EJS (Kirkwood SoC) + Tom Warren <[email protected]> harmony Tegra2 (ARM7 & A9 Dual Core)
board/buffalo/lsxl/Makefile new file +50 -0
50 lines not on this page · raw message
board/buffalo/lsxl/kwbimage-lschl.cfg new file +229 -0
229 lines not on this page · raw message
board/buffalo/lsxl/kwbimage-lsxhl.cfg new file +229 -0
229 lines not on this page · raw message
board/buffalo/lsxl/lsxl.c new file +302 -0
302 lines not on this page · raw message
board/buffalo/lsxl/lsxl.h new file +75 -0
75 lines not on this page · raw message
boards.cfg +2 -0
@@ -137,6 +137,8 @@ hawkboard_uart arm arm926ejs da8xxevm davinci enbw_cmc arm arm926ejs enbw_cmc enbw davinci calimain arm arm926ejs calimain omicron davinci dns325 arm arm926ejs - d-link kirkwood +lschlv2 arm arm926ejs lsxl buffalo kirkwood lsxl:LSCHLV2 +lsxhl arm arm926ejs lsxl buffalo kirkwood lsxl:LSXHL km_kirkwood arm arm926ejs km_arm keymile kirkwood km_kirkwood:KM_DISABLE_PCI km_kirkwood_pci arm arm926ejs km_arm keymile kirkwood km_kirkwood:KM_RECONFIG_XLX mgcoge3un arm arm926ejs km_arm keymile kirkwood
include/configs/lsxl.h new file +182 -0
182 lines not on this page · raw message
Signature · 1 line
1.7.2.5
PATCH v5 5/5 Kirkwood: add lschlv2 and lsxhl board support
[sorry my first mail didn't have 'Kirkwood' in the subject]
Quoted 2 lines
This patch adds support for both the Linkstation Live (LS-CHLv2) and Linkstation Pro (LS-XHL) by Buffalo.
Am Freitag 11 Mai 2012, 18:51:09 schrieb Prafulla Wadaskar:
Quoted 1 line
I will check on this patch in this week end since I am travelling.
Hi Prafulla,
did you find the time to review the arm/board specific patches in this series
(http://patchwork.ozlabs.org/patch/158627/)?
Signature · 1 line
Michael
PATCH v5 5/5 Kirkwood: add lschlv2 and lsxhl board support
Quoted 16 lines
-----Original Message-----
From: Michael Walle [mailto:michael at walle.cc]
Sent: 23 May 2012 02:28
To: u-boot at lists.denx.de
Cc: Prafulla Wadaskar
Subject: Re: [PATCH v5 5/5] Kirkwood: add lschlv2 and lsxhl board support[sorry my first mail didn't have 'Kirkwood' in the subject]
Quoted 2 lines
This patch adds support for both the Linkstation Live (LS-CHLv2) and Linkstation Pro (LS-XHL) by Buffalo.
Quoted 8 lines
Am Freitag 11 Mai 2012, 18:51:09 schrieb Prafulla Wadaskar:
Quoted 1 line
I will check on this patch in this week end since I am travelling.
Hi Prafulla,
did you find the time to review the arm/board specific patches in this series
(http://patchwork.ozlabs.org/patch/158627/)?
I will do it on this weekend
Regards..
Prafulla . . .
PATCH v5 5/5 Kirkwood: add lschlv2 and lsxhl board support
In reply to Michael WalleQuoted 15 lines
-----Original Message-----
From: Michael Walle [mailto:michael at walle.cc]
Sent: 12 May 2012 04:21
To: u-boot at lists.denx.de
Cc: Prafulla Wadaskar; Wolfgang Denk; Mike Frysinger; Joe Hershberger; Michael Walle
Subject: [PATCH v5 5/5] Kirkwood: add lschlv2 and lsxhl board supportThis patch adds support for both the Linkstation Live (LS-CHLv2) and Linkstation Pro (LS-XHL) by Buffalo.
Signed-off-by Michael Walle <[email protected]>Cc Prafulla Wadaskar <[email protected]>
Quoted 7 lines
--- MAINTAINERS | 5 + board/buffalo/lsxl/Makefile | 50 ++++++ board/buffalo/lsxl/kwbimage-lschl.cfg | 229 +++++++++++++++++++++++++ board/buffalo/lsxl/kwbimage-lsxhl.cfg | 229+++++++++++++++++++++++++
BTW: These two file looks similar, what is the difference?
Quoted 815 lines
board/buffalo/lsxl/lsxl.c | 302 +++++++++++++++++++++++++++++++++ board/buffalo/lsxl/lsxl.h | 75 ++++++++ boards.cfg | 2 + include/configs/lsxl.h | 182 ++++++++++++++++++++ 8 files changed, 1074 insertions(+), 0 deletions(-) create mode 100644 board/buffalo/lsxl/Makefile create mode 100644 board/buffalo/lsxl/kwbimage-lschl.cfg create mode 100644 board/buffalo/lsxl/kwbimage-lsxhl.cfg create mode 100644 board/buffalo/lsxl/lsxl.c create mode 100644 board/buffalo/lsxl/lsxl.h create mode 100644 include/configs/lsxl.hMAINTAINERS +5
-0@@ -894,6 +894,11 @@ Prafulla Wadaskar <[email protected]> rd6281a ARM926EJS (Kirkwood SoC) sheevaplug ARM926EJS (Kirkwood SoC) +Michael Walle <[email protected]> + + lschlv2 ARM926EJS (Kirkwood SoC) + lsxhl ARM926EJS (Kirkwood SoC) + Tom Warren <[email protected]> harmony Tegra2 (ARM7 & A9 Dual Core)board/buffalo/lsxl/Makefile new file +43
-0@@ -0,0 +1,50 @@ +# +# Copyright (c) 2012 Michael Walle +# Michael Walle <[email protected]> +# +# See file CREDITS for list of people who contributed to this +# project. +# +# This program is free software; you can redistribute it and/or +# modify it under the terms of the GNU General Public License as +# published by the Free Software Foundation; either version 2 of +# the License, or (at your option) any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program; if not, write to the Free Software +# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, +# MA 02110-1301 USA +# + +include $(TOPDIR)/config.mk + +LIB = $(obj)lib$(BOARD).o + +COBJS := lsxl.o + +SRCS := $(SOBJS:.o=.S) $(COBJS:.o=.c) +OBJS := $(addprefix $(obj),$(COBJS)) +SOBJS := $(addprefix $(obj),$(SOBJS)) + +$(LIB): $(obj).depend $(OBJS) $(SOBJS) + $(call cmd_link_o_target, $(OBJS) $(SOBJS)) + +clean: + rm -f $(SOBJS) $(OBJS) + +distclean: clean + rm -f $(LIB) core *.bak .depend + +#########################################################################
board/buffalo/lsxl/Makefile new file +7
-0+ +# defines $(obj).depend target +include $(SRCTREE)/rules.mk + +sinclude $(obj).depend + +#########################################################################
board/buffalo/lsxl/kwbimage-lschl.cfg new file +38
-0@@ -0,0 +1,229 @@ +# +# Copyright (c) 2012 Michael Walle +# Michael Walle <[email protected]> +# +# See file CREDITS for list of people who contributed to this +# project. +# +# This program is free software; you can redistribute it and/or +# modify it under the terms of the GNU General Public License as +# published by the Free Software Foundation; either version 2 of +# the License, or (at your option) any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program; if not, write to the Free Software +# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, +# MA 02110-1301 USA +# +# Refer docs/README.kwimage for more details about how-to configure +# and create kirkwood boot image +# + +# Boot Media configurations +BOOT_FROM spi + +# SOC registers configuration using bootrom header extension +# Maximum KWBIMAGE_MAX_CONFIG configurations allowed + +# Configure RGMII-0/1 interface pad voltage to 1.8V +DATA 0xFFD100E0 0x1B1B1B9B + +# L2 RAM Timing 0 +DATA 0xFFD20134 0xBBBBBBBB +# not further specified in HW manual, timing taken from originalvendor port
board/buffalo/lsxl/kwbimage-lschl.cfg new file +4
-0+ +# L2 RAM Timing 1 +DATA 0xFFD20138 0x00BBBBBB +# not further specified in HW manual, timing taken from originalvendor port
board/buffalo/lsxl/kwbimage-lschl.cfg new file +15
-0+ +# DDR Configuration register +DATA 0xFFD01400 0x43000618 +# bit13-0: 0x618, 1560 DDR2 clks refresh rate +# bit23-14: 0 required +# bit24: 1, enable exit self refresh mode on DDR access +# bit25: 1 required +# bit29-26: 0 required +# bit31-30: 0b01 required + +# DDR Controller Control Low +DATA 0xFFD01404 0x39543000 +# bit3-0: 0 required +# bit4: 0, addr/cmd in same cycle +# bit5: 0, clk is driven during self refresh, we don't care forAPX
board/buffalo/lsxl/kwbimage-lschl.cfg new file +9
-0+# bit6: 0, use recommended falling edge of clk for addr/cmd +# bit11-7: 0 required +# bit12: 1 required +# bit13: 1 required +# bit14: 0, input buffer always powered up +# bit17-15: 0 required +# bit18: 1, cpu lock transaction enabled +# bit19: 0 required +# bit23-20: 5, recommended value for CL=5 and STARTBURST_DEL disabledbit31=0
+# bit27-24: 9, CL+4, STARTBURST sample stages, for freqs 400MHz, unbuffered DIMMboard/buffalo/lsxl/kwbimage-lschl.cfg new file +80
-0+# bit30-28: 3 required +# bit31: 0, no additional STARTBURST delay + +# DDR Timing (Low) +DATA 0xFFD01408 0x3302444F +# bit3-0: 0xf, 16 cycle tRAS (tRAS[3-0]) +# bit7-4: 4, 5 cycle tRCD +# bit11-8: 4, 5 cyle tRP +# bit15-12: 4, 5 cyle tWR +# bit19-16: 2, 3 cyle tWTR +# bit20: 0, 16 cycle tRAS (tRAS[4]) +# bit23-21: 0 required +# bit27-24: 3, 4 cycle tRRD +# bit31-28: 3, 4 cyle tRTP + +# DDR Timing (High) +DATA 0xFFD0140C 0x00000823 +# bit6-0: 0x23, 35 cycle tRFC +# bit8-7: 0, 1 cycle tR2R +# bit10-9: 0, 1 cyle tR2W +# bit12-11: 1, 2 cylce tW2W +# bit31-13: 0 required + +# DDR Address Control +DATA 0xFFD01410 0x00000009 +# bit1-0: 1, Cs0width=x16 +# bit3-2: 2, Cs0size=512Mbit +# bit5-4: 0, Cs1width=nonexistent +# bit7-6: 0, Cs1size=nonexistent +# bit9-8: 0, Cs2width=nonexistent +# bit11-10: 0, Cs2size=nonexistent +# bit13-12: 0, Cs3width=nonexistent +# bit15-14: 0, Cs3size=nonexistent +# bit16: 0, Cs0AddrSel +# bit17: 0, Cs1AddrSel +# bit18: 0, Cs2AddrSel +# bit19: 0, Cs3AddrSel +# bit31-20: 0 required + +# DDR Open Pages Control +DATA 0xFFD01414 0x00000000 +# bit0: 0, OPEn=OpenPage enabled +# bit31-1: 0 required + +# DDR Operation +DATA 0xFFD01418 0x00000000 +# bit3-0: 0, Cmd=Normal SDRAM Mode +# bit31-4: 0 required + +# DDR Mode +DATA 0xFFD0141C 0x00000652 +# bit2-0: 2, Burst Length (2 required) +# bit3: 0, Burst Type (0 required) +# bit6-4: 5, CAS Latency (CL) 5 +# bit7: 0, (Test Mode) Normal operation +# bit8: 0, (Reset DLL) Normal operation +# bit11-9: 3, Write recovery for auto-precharge (3 required) +# bit12: 0, Fast Active power down exit time (0 required) +# bit31-13: 0 required + +# DDR Extended Mode +DATA 0xFFD01420 0x00000042 +# bit0: 0, DRAM DLL enabled +# bit1: 1, DRAM drive strength reduced +# bit2: 0, ODT control Rtt[0] (Rtt=2, 150 ohm termination) +# bit5-3: 0 required +# bit6: 1, ODT control Rtt[1] (Rtt=2, 150 ohm termination) +# bit9-7: 0 required +# bit10: 0, differential DQS enabled +# bit11: 0 required +# bit12: 0, DRAM output buffer enabled +# bit31-13: 0 required + +# DDR Controller Control High +DATA 0xFFD01424 0x0000F17F +# bit2-0: 0x7 required +# bit3: 1, MBUS Burst Chop disabled +# bit6-4: 0x7 required +# bit7: 0 required (???) +# bit8: 1, add writepath sample stage, must be 1 for DDR freq >=300MHz
board/buffalo/lsxl/kwbimage-lschl.cfg new file +10
-0+# bit9: 0, no half clock cycle addition to dataout +# bit10: 0, 1/4 clock cycle skew enabled for addr/ctl signals +# bit11: 0, 1/4 clock cycle skew disabled for write mesh +# bit15-12: 0xf required +# bit31-16: 0 required + +# DDR2 ODT Read Timing (default values) +DATA 0xFFD01428 0x00085520 +# bit3-0: 0 required +# bit7-4: 2, 2 cycles from read command to assertion of M_ODTsignal
+# bit11-8: 5, 5 cycles from read command to de-assertion ofM_ODTsignal
+# bit15-12: 5, 5 cycles from read command to assertion of internal ODT signal
+# bit19-16: 8, 8 cycles from read command to de-assertion of internal ODT signalboard/buffalo/lsxl/kwbimage-lschl.cfg new file +5
-0+# bit31-20: 0 required + +# DDR2 ODT Write Timing (default values) +DATA 0xFFD0147C 0x00008552 +# bit3-0: 2, 2 cycles from write comand to assertion of M_ODTsignal
+# bit7-4: 5, 5 cycles from write command to de-assertion ofM_ODT
signal
+# bit15-12: 5, 5 cycles from write command to assertion of internal ODT signal
+# bit19-16: 8, 8 cycles from write command to de-assertion of internal ODT signalboard/buffalo/lsxl/kwbimage-lschl.cfg new file +29
-0+# bit31-16: 0 required + +# CS[0]n Base address +DATA 0xFFD01500 0x00000000 +# at 0x0 + +# CS[0]n Size +DATA 0xFFD01504 0x03FFFFF1 +# bit0: 1, Window enabled +# bit1: 0, Write Protect disabled +# bit3-2: 0x0, CS0 hit selected +# bit23-4: 0xfffff required +# bit31-24: 0x03, Size (i.e. 64MB) + +# CS[1]n Size +DATA 0xFFD0150C 0x00000000 +# window disabled + +# CS[2]n Size +DATA 0xFFD01514 0x00000000 +# window disabled + +# CS[3]n Size +DATA 0xFFD0151C 0x00000000 +# window disabled + +# DDR ODT Control (Low) +DATA 0xFFD01494 0x003C0000 +# bit3-0: 0b0000, (read) M_ODT[0] is not asserted during read fromDRAM
+# bit7-4: 0b0000, (read) M_ODT[1] is not asserted during read from
DRAMboard/buffalo/lsxl/kwbimage-lschl.cfg new file +2
-0+# bit15-8: 0 required +# bit19-16: 0b1100, (write) M_ODT[0] is asserted during write to DRAMCS2, CS3
+# bit23-20: 0b0011, (write) M_ODT[1] is asserted during write to DRAM CS0, CS1board/buffalo/lsxl/kwbimage-lschl.cfg new file +5
-0+# bit31-24: 0 required + +# DDR ODT Control (High) +DATA 0xFFD01498 0x00000000 +# bit1-0: 0, M_ODT[0] assertion is controlled by ODT Control Lowregister
+# bit3-2: 0, M_ODT[1] assertion is controlled by ODT Control Low
registerboard/buffalo/lsxl/kwbimage-lschl.cfg new file +5
-0+# bit31-4 0 required + +# CPU ODT Control +DATA 0xFFD0149C 0x0000E80F +# bit3-0: 0b1111, internal ODT is asserted during read from DRAMbank 0-3
+# bit7-4: 0b0000, internal ODT is not asserted during write to DRAM
bank 0-3board/buffalo/lsxl/kwbimage-lschl.cfg new file +16
-0+# bit9-8: 0, Internal ODT assertion is controlled by fiels +# bit11-10: 2, M_DQ, M_DM, and M_DQS I/O buffer ODT 75 ohm +# bit13-12: 2, M_STARTBURST_IN I/O buffer ODT 75 ohm +# bit14: 1, M_STARTBURST_IN ODT enabled +# bit15: 1, DDR IO ODT Unit: Drive ODT calibration values +# bit20-16: 0, Pad N channel driving strength for ODT +# bit25-21: 0, Pad P channel driving strength for ODT +# bit31-26: 0 required + +# DDR Initialization Control +DATA 0xFFD01480 0x00000001 +# bit0: 1, enable DDR init upon this register write +# bit31-1: 0, required + +# End of Header extension +DATA 0x0 0x0board/buffalo/lsxl/kwbimage-lsxhl.cfg new file +38
-0@@ -0,0 +1,229 @@ +# +# Copyright (c) 2012 Michael Walle +# Michael Walle <[email protected]> +# +# See file CREDITS for list of people who contributed to this +# project. +# +# This program is free software; you can redistribute it and/or +# modify it under the terms of the GNU General Public License as +# published by the Free Software Foundation; either version 2 of +# the License, or (at your option) any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program; if not, write to the Free Software +# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, +# MA 02110-1301 USA +# +# Refer docs/README.kwimage for more details about how-to configure +# and create kirkwood boot image +# + +# Boot Media configurations +BOOT_FROM spi + +# SOC registers configuration using bootrom header extension +# Maximum KWBIMAGE_MAX_CONFIG configurations allowed + +# Configure RGMII-0/1 interface pad voltage to 1.8V +DATA 0xFFD100E0 0x1B1B9B9B + +# L2 RAM Timing 0 +DATA 0xFFD20134 0xBBBBBBBB +# not further specified in HW manual, timing taken from originalvendor port
board/buffalo/lsxl/kwbimage-lsxhl.cfg new file +4
-0+ +# L2 RAM Timing 1 +DATA 0xFFD20138 0x00BBBBBB +# not further specified in HW manual, timing taken from originalvendor port
board/buffalo/lsxl/kwbimage-lsxhl.cfg new file +15
-0+ +# DDR Configuration register +DATA 0xFFD01400 0x43000618 +# bit13-0: 0x618, 1560 DDR2 clks refresh rate +# bit23-14: 0 required +# bit24: 1, enable exit self refresh mode on DDR access +# bit25: 1 required +# bit29-26: 0 required +# bit31-30: 0b01 required + +# DDR Controller Control Low +DATA 0xFFD01404 0x39543010 +# bit3-0: 0 required +# bit4: 1, T2 mode, addr/cmd are driven for two cycles +# bit5: 0, clk is driven during self refresh, we don't care forAPX
board/buffalo/lsxl/kwbimage-lsxhl.cfg new file +9
-0+# bit6: 0, use recommended falling edge of clk for addr/cmd +# bit11-7: 0 required +# bit12: 1 required +# bit13: 1 required +# bit14: 0, input buffer always powered up +# bit17-15: 0 required +# bit18: 1, cpu lock transaction enabled +# bit19: 0 required +# bit23-20: 5, recommended value for CL=5 and STARTBURST_DEL disabledbit31=0
+# bit27-24: 9, CL+4, STARTBURST sample stages, for freqs 400MHz, unbuffered DIMMboard/buffalo/lsxl/kwbimage-lsxhl.cfg new file +80
-0+# bit30-28: 3 required +# bit31: 0, no additional STARTBURST delay + +# DDR Timing (Low) +DATA 0xFFD01408 0x22125441 +# bit3-0: 0x1, 18 cycle tRAS (tRAS[3-0]) +# bit7-4: 4, 5 cycle tRCD +# bit11-8: 4, 5 cyle tRP +# bit15-12: 5, 6 cyle tWR +# bit19-16: 2, 3 cyle tWTR +# bit20: 1, 18 cycle tRAS (tRAS[4]) +# bit23-21: 0 required +# bit27-24: 2, 3 cycle tRRD +# bit31-28: 2, 3 cyle tRTP + +# DDR Timing (High) +DATA 0xFFD0140C 0x00000832 +# bit6-0: 0x32, 50 cycle tRFC +# bit8-7: 0, 1 cycle tR2R +# bit10-9: 0, 1 cyle tR2W +# bit12-11: 1, 2 cylce tW2W +# bit31-13: 0 required + +# DDR Address Control +DATA 0xFFD01410 0x0000000C +# bit1-0: 0, Cs0width=x8 +# bit3-2: 3, Cs0size=1Gbit +# bit5-4: 0, Cs1width=nonexistent +# bit7-6: 0, Cs1size=nonexistent +# bit9-8: 0, Cs2width=nonexistent +# bit11-10: 0, Cs2size=nonexistent +# bit13-12: 0, Cs3width=nonexistent +# bit15-14: 0, Cs3size=nonexistent +# bit16: 0, Cs0AddrSel +# bit17: 0, Cs1AddrSel +# bit18: 0, Cs2AddrSel +# bit19: 0, Cs3AddrSel +# bit31-20: 0 required + +# DDR Open Pages Control +DATA 0xFFD01414 0x00000000 +# bit0: 0, OPEn=OpenPage enabled +# bit31-1: 0 required + +# DDR Operation +DATA 0xFFD01418 0x00000000 +# bit3-0: 0, Cmd=Normal SDRAM Mode +# bit31-4: 0 required + +# DDR Mode +DATA 0xFFD0141C 0x00000652 +# bit2-0: 2, Burst Length (2 required) +# bit3: 0, Burst Type (0 required) +# bit6-4: 5, CAS Latency (CL) 5 +# bit7: 0, (Test Mode) Normal operation +# bit8: 0, (Reset DLL) Normal operation +# bit11-9: 3, Write recovery for auto-precharge (3 required) +# bit12: 0, Fast Active power down exit time (0 required) +# bit31-13: 0 required + +# DDR Extended Mode +DATA 0xFFD01420 0x00000006 +# bit0: 0, DRAM DLL enabled +# bit1: 1, DRAM drive strength reduced +# bit2: 1, ODT control Rtt[0] (Rtt=1, 75 ohm termination) +# bit5-3: 0 required +# bit6: 0, ODT control Rtt[1] (Rtt=1, 75 ohm termination) +# bit9-7: 0 required +# bit10: 0, differential DQS enabled +# bit11: 0 required +# bit12: 0, DRAM output buffer enabled +# bit31-13: 0 required + +# DDR Controller Control High +DATA 0xFFD01424 0x0000F17F +# bit2-0: 0x7 required +# bit3: 1, MBUS Burst Chop disabled +# bit6-4: 0x7 required +# bit7: 0 required (???) +# bit8: 1, add writepath sample stage, must be 1 for DDR freq >=300MHz
board/buffalo/lsxl/kwbimage-lsxhl.cfg new file +10
-0+# bit9: 0, no half clock cycle addition to dataout +# bit10: 0, 1/4 clock cycle skew enabled for addr/ctl signals +# bit11: 0, 1/4 clock cycle skew disabled for write mesh +# bit15-12: 0xf required +# bit31-16: 0 required + +# DDR2 ODT Read Timing (default values) +DATA 0xFFD01428 0x00085520 +# bit3-0: 0 required +# bit7-4: 2, 2 cycles from read command to assertion of M_ODTsignal
+# bit11-8: 5, 5 cycles from read command to de-assertion ofM_ODTsignal
+# bit15-12: 5, 5 cycles from read command to assertion of internal ODT signal
+# bit19-16: 8, 8 cycles from read command to de-assertion of internal ODT signalboard/buffalo/lsxl/kwbimage-lsxhl.cfg new file +5
-0+# bit31-20: 0 required + +# DDR2 ODT Write Timing (default values) +DATA 0xFFD0147C 0x00008552 +# bit3-0: 2, 2 cycles from write comand to assertion of M_ODTsignal
+# bit7-4: 5, 5 cycles from write command to de-assertion ofM_ODT
signal
+# bit15-12: 5, 5 cycles from write command to assertion of internal ODT signal
+# bit19-16: 8, 8 cycles from write command to de-assertion of internal ODT signalboard/buffalo/lsxl/kwbimage-lsxhl.cfg new file +29
-0+# bit31-16: 0 required + +# CS[0]n Base address +DATA 0xFFD01500 0x00000000 +# at 0x0 + +# CS[0]n Size +DATA 0xFFD01504 0x0FFFFFF1 +# bit0: 1, Window enabled +# bit1: 0, Write Protect disabled +# bit3-2: 0x0, CS0 hit selected +# bit23-4: 0xfffff required +# bit31-24: 0x0f, Size (i.e. 256MB) + +# CS[1]n Size +DATA 0xFFD0150C 0x00000000 +# window disabled + +# CS[2]n Size +DATA 0xFFD01514 0x00000000 +# window disabled + +# CS[3]n Size +DATA 0xFFD0151C 0x00000000 +# window disabled + +# DDR ODT Control (Low) +DATA 0xFFD01494 0x00010000 +# bit3-0: 0b0000, (read) M_ODT[0] is not asserted during read fromDRAM
+# bit7-4: 0b0000, (read) M_ODT[1] is not asserted during read from
DRAMboard/buffalo/lsxl/kwbimage-lsxhl.cfg new file +2
-0+# bit15-8: 0 required +# bit19-16: 0b0001, (write) M_ODT[0] is asserted during write to DRAMCS0
+# bit23-20: 0b0000, (write) M_ODT[1] is not asserted during write to DRAMboard/buffalo/lsxl/kwbimage-lsxhl.cfg new file +5
-0+# bit31-24: 0 required + +# DDR ODT Control (High) +DATA 0xFFD01498 0x00000000 +# bit1-0: 0, M_ODT[0] assertion is controlled by ODT Control Lowregister
+# bit3-2: 0, M_ODT[1] assertion is controlled by ODT Control Low
registerboard/buffalo/lsxl/kwbimage-lsxhl.cfg new file +5
-0+# bit31-4 0 required + +# CPU ODT Control +DATA 0xFFD0149C 0x0000E80F +# bit3-0: 0b1111, internal ODT is asserted during read from DRAMbank 0-3
+# bit7-4: 0b0000, internal ODT is not asserted during write to DRAM
bank 0-3board/buffalo/lsxl/kwbimage-lsxhl.cfg new file +16
-0+# bit9-8: 0, Internal ODT assertion is controlled by fiels +# bit11-10: 2, M_DQ, M_DM, and M_DQS I/O buffer ODT 75 ohm +# bit13-12: 2, M_STARTBURST_IN I/O buffer ODT 75 ohm +# bit14: 1, M_STARTBURST_IN ODT enabled +# bit15: 1, DDR IO ODT Unit: Drive ODT calibration values +# bit20-16: 0, Pad N channel driving strength for ODT +# bit25-21: 0, Pad P channel driving strength for ODT +# bit31-26: 0 required + +# DDR Initialization Control +DATA 0xFFD01480 0x00000001 +# bit0: 1, enable DDR init upon this register write +# bit31-1: 0, required + +# End of Header extension +DATA 0x0 0x0board/buffalo/lsxl/lsxl.c new file +43
-0@@ -0,0 +1,302 @@ +/* +* Copyright (c) 2012 Michael Walle +* Michael Walle <[email protected]> +* +* Based on sheevaplug/sheevaplug.c by +* Marvell Semiconductor <www.marvell.com> +* +* See file CREDITS for list of people who contributed to this +* project. +* +* This program is free software; you can redistribute it and/or +* modify it under the terms of the GNU General Public License as +* published by the Free Software Foundation; either version 2 of +* the License, or (at your option) any later version. +* +* This program is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +* GNU General Public License for more details. +* +* You should have received a copy of the GNU General Public License +* along with this program; if not, write to the Free Software +* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, +* MA 02110-1301 USA +*/ + +#include <common.h> +#include <net.h> +#include <malloc.h> +#include <netdev.h> +#include <miiphy.h> +#include <asm/arch/kirkwood.h> +#include <asm/arch/cpu.h> +#include <asm/arch/mpp.h> +#include <asm/arch/gpio.h> +#include <spi_flash.h> + +#include "lsxl.h" + +/* +* Rescue mode +* +* Selected by holding the push button for 3 seconds, while powering on
+ * the device.
+ *
+ * These linkstations don't have a (populated) serial port. There is no
+ * way to access an (unmodified) board other than using the netconsole. If
+ * you want to recover from a bad environment setting or an empty environment,
+ * you can do this only with a working network connection. Therefore, a random
+ * ethernet address is generated if none is set and a DHCP request is sent.
+ * After a successful DHCP response is received, the network settings are
+ * configured and the ncip parameter is set to the serverip. Eg. for a working
+ * resuce mode, you should set 'next-server' to the host where the netconsole
+ * client is started.
+ * Additionally, the bootsource is set to 'cli'.
+ */board/buffalo/lsxl/lsxl.c new file +12
-0+ +#ifndef CONFIG_ENV_OVERWRITE +# error "You need to set CONFIG_ENV_OVERWRITE" +#endif + +DECLARE_GLOBAL_DATA_PTR ; + +int board_early_init_f (void ) +{ +/* +* default gpio configuration +* There are maximum 64 gpios controlled through 2 sets of registers
board/buffalo/lsxl/lsxl.c new file +78
-0+* the below configuration configures mainly initial LED status +*/ +kw_config_gpio (LSXL_OE_VAL_LOW , +LSXL_OE_VAL_HIGH , +LSXL_OE_LOW ,LSXL_OE_HIGH ); + +/* +* Multi-Purpose Pins Functionality configuration +* These strappings are taken from the original vendor uboot port. +*/ +u32 kwmpp_config[] = { + MPP0_SPI_SCn, +MPP1_SPI_MOSI , +MPP2_SPI_SCK , +MPP3_SPI_MISO , +MPP4_UART0_RXD , +MPP5_UART0_TXD , + MPP6_SYSRST_OUTn, +MPP7_GPO , +MPP8_GPIO , +MPP9_GPIO , +MPP10_GPO ,/* HDD power */ +MPP11_GPIO ,/* USB Vbus enable */ +MPP12_SD_CLK , +MPP13_SD_CMD , +MPP14_SD_D0 , +MPP15_SD_D1 , +MPP16_SD_D2 , +MPP17_SD_D3 , +MPP18_GPO ,/* fan speed high */ +MPP19_GPO ,/* fan speed low */ +MPP20_GE1_0 , +MPP21_GE1_1 , +MPP22_GE1_2 , +MPP23_GE1_3 , +MPP24_GE1_4 , +MPP25_GE1_5 , +MPP26_GE1_6 , +MPP27_GE1_7 , +MPP28_GPIO , +MPP29_GPIO , +MPP30_GE1_10 , +MPP31_GE1_11 , +MPP32_GE1_12 , +MPP33_GE1_13 , +MPP34_GPIO , +MPP35_GPIO , +MPP36_GPIO ,/* function LED */ +MPP37_GPIO ,/* alarm LED */ +MPP38_GPIO ,/* info LED */ +MPP39_GPIO ,/* power LED */ +MPP40_GPIO ,/* fan alarm */ +MPP41_GPIO ,/* funtion button */ +MPP42_GPIO ,/* power switch */ +MPP43_GPIO ,/* power auto switch */ +MPP44_GPIO , +MPP45_GPIO , +MPP46_GPIO , +MPP47_GPIO , +MPP48_GPIO ,/* function red LED */ +MPP49_GPIO , +0 + }; + +kirkwood_mpp_conf (kwmpp_config); + +return 0 ; +} + +#define LED_OFF 0 +#define LED_ALARM_ON 1 +#define LED_ALARM_BLINKING 2 +#define LED_POWER_ON 3 +#define LED_POWER_BLINKING 4 +#define LED_INFO_ON 5 +#define LED_INFO_BLINKING 6 + +static void __set_led(int blink_alarm, int blink_info, int
blink_power,board/buffalo/lsxl/lsxl.c new file +49
-0+ int value_alarm, int value_info, int value_power) +{ +kw_gpio_set_blink (GPIO_ALARM_LED ,blink_alarm ); +kw_gpio_set_blink (GPIO_INFO_LED , blink_info); +kw_gpio_set_blink (GPIO_POWER_LED , blink_power); +kw_gpio_set_value (GPIO_ALARM_LED , value_alarm); +kw_gpio_set_value (GPIO_INFO_LED , value_info); +kw_gpio_set_value (GPIO_POWER_LED , value_power); +} + +static void set_led (int state) +{ +switch (state) { +case LED_OFF : +__set_led (0 ,0 ,0 ,0 ,0 ,0 ); +break ; +case LED_ALARM_ON : +__set_led (0 ,0 ,0 ,0 ,1 ,1 ); +break ; +case LED_ALARM_BLINKING : +__set_led (1 ,0 ,0 ,1 ,1 ,1 ); +break ; +case LED_INFO_ON : +__set_led (0 ,0 ,0 ,1 ,0 ,1 ); +break ; +case LED_INFO_BLINKING : +__set_led (0 ,1 ,0 ,1 ,1 ,1 ); +break ; +case LED_POWER_ON : +__set_led (0 ,0 ,0 ,1 ,1 ,0 ); +break ; +case LED_POWER_BLINKING : +__set_led (0 ,0 ,1 ,1 ,1 ,1 ); +break ; + } +} + +int board_init (void ) +{ +/* address of boot parameters */ + gd->bd->bi_boot_params =kw_sdram_bar (0 ) +0x100 ; + +set_led (LED_POWER_BLINKING ); + +return 0 ; +} + +#ifdef CONFIG_MISC_INIT_R +void check_enetaddr (void )
board/buffalo/lsxl/lsxl.c line 205
Quoted 10 lines
board/buffalo/lsxl/lsxl.c new file +10
-0+{ +uchar enetaddr[6 ]; + +if (!eth_getenv_enetaddr_by_index ("eth" ,0 , enetaddr)) { +/* signal unset/invalid ethaddr to user */ +set_led (LED_INFO_BLINKING ); + } +} + +static void erase_environment (void )
Why do you need this function?
Quoted 43 lines
board/buffalo/lsxl/lsxl.c new file +43
-0+{ +struct spi_flash *flash; + +printf ("Erasing environment..\n" ); + flash =spi_flash_probe (0 ,0 ,1000000 ,SPI_MODE_3 ); +if (!flash) { +printf ("Erasing flash failed\n" ); +return ; + } + +spi_flash_erase (flash,CONFIG_ENV_OFFSET ,CONFIG_ENV_SIZE ); +spi_flash_free (flash); +do_reset (NULL ,0 ,0 ,NULL ); +} + +static void rescue_mode (void ) +{ +uchar enetaddr[6 ]; + +printf ("Entering rescue mode..\n" ); +if (!eth_getenv_enetaddr_by_index ("eth" ,0 , enetaddr)) { +eth_random_enetaddr (enetaddr); +if (eth_setenv_enetaddr_by_index ("eth" ,0 , enetaddr)) { +printf ("Failed to set ethernet address\n" ); +set_led (LED_ALARM_BLINKING ); +return ; + } + } +setenv ("bootsource" ,"rescue" ); +} + +static void check_push_button (void ) +{ +int i =0 ; + +while (!kw_gpio_get_value (GPIO_FUNC_BUTTON )) { +udelay (100000 ); + i++; + +if (i ==10 ) +set_led (LED_INFO_ON ); + +if (i >=100 ) {
board/buffalo/lsxl/lsxl.c line 258
Quoted 10 lines
board/buffalo/lsxl/lsxl.c new file +10
-0+set_led (LED_INFO_BLINKING ); +break ; + } + } + +if (i >=100 ) +erase_environment (); +else if (i >=10 ) +rescue_mode (); +}
You can overload this functionality to Reset key, you don't need any supporting code for the same, you can handle it through additional environment variables. For more info check for "SYSTSTn Duration counter support" in Kirkwood/cpu.c
board/buffalo/lsxl/lsxl.c line 262
Quoted 4 lines
board/buffalo/lsxl/lsxl.c new file +4
-0+ +int misc_init_r (void ) +{ +check_enetaddr ();
This need to be ifdefed with CONFIG_MISC_INIT_R
Quoted 190 lines
board/buffalo/lsxl/lsxl.c new file +13
-0+check_push_button (); + +return 0 ; +} +#endif + +#ifdef CONFIG_SHOW_BOOT_PROGRESS +void show_boot_progress (int progress) +{ +if (progress >0 ) +return ; + +/* this is not an error, eg. bootp with autoload=no will trigger this */
board/buffalo/lsxl/lsxl.c new file +27
-0+if (progress == -BOOTSTAGE_ID_NET_LOADED ) +return ; + + set_led(LED_ALARM_BLINKING); +} +#endif + +#ifdef CONFIG_RESET_PHY_R +/* Configure and enable MV88E1118 PHY */ +void reset_phy (void ) +{ +u16 devadr; +char *name ="egiga1" ; + +if (miiphy_set_current_dev (name)) +return ; + +/* command to read PHY dev address */ +if (miiphy_read (name,0xEE ,0xEE , (u16 *) &devadr)) { +printf ("Err..%s could not read PHY dev address\n" , __func__); +return ; + } + +/* reset the phy */ +miiphy_reset (name, devadr); +} +#endif /* CONFIG_RESET_PHY_R */ board/buffalo/lsxl/lsxl.h new file +75
-0@@ -0,0 +1,75 @@ +/* +* Copyright (c) 2012 Michael Walle +* Michael Walle <[email protected]> +* +* See file CREDITS for list of people who contributed to this +* project. +* +* This program is free software; you can redistribute it and/or +* modify it under the terms of the GNU General Public License as +* published by the Free Software Foundation; either version 2 of +* the License, or (at your option) any later version. +* +* This program is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +* GNU General Public License for more details. +* +* You should have received a copy of the GNU General Public License +* along with this program; if not, write to the Free Software +* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, +* MA 02110-1301 USA +*/ + +#ifndef __LSXL_H +#define __LSXL_H + +#define GPIO_HDD_POWER 10 +#define GPIO_USB_VBUS 11 +#define GPIO_FAN_HIGH 18 +#define GPIO_FAN_LOW 19 +#define GPIO_FUNC_LED 36 +#define GPIO_ALARM_LED 37 +#define GPIO_INFO_LED 38 +#define GPIO_POWER_LED 39 +#define GPIO_FAN_LOCK 40 +#define GPIO_FUNC_BUTTON 41 +#define GPIO_POWER_SWITCH 42 +#define GPIO_POWER_AUTO_SWITCH 43 +#define GPIO_FUNC_RED_LED 48 + +#define _BIT (x) (1<<(x)) + +#define LSXL_OE_LOW (~(_BIT(GPIO_HDD_POWER) \ + | _BIT(GPIO_USB_VBUS) \ + | _BIT(GPIO_FAN_HIGH) \ + | _BIT(GPIO_FAN_LOW))) + +#define LSXL_OE_HIGH (~(_BIT(GPIO_FUNC_LED - 32) \ + | _BIT(GPIO_ALARM_LED - 32) \ + | _BIT(GPIO_INFO_LED - 32) \ + | _BIT(GPIO_POWER_LED - 32) \ + | _BIT(GPIO_FUNC_RED_LED - 32))) + +#define LSXL_OE_VAL_LOW (_BIT(GPIO_HDD_POWER) \ + | _BIT(GPIO_USB_VBUS)) + +#define LSXL_OE_VAL_HIGH (_BIT(GPIO_FUNC_LED - 32) \ + | _BIT(GPIO_ALARM_LED - 32) \ + | _BIT(GPIO_INFO_LED - 32) \ + | _BIT(GPIO_POWER_LED - 32) \ + | _BIT(GPIO_FUNC_RED_LED - 32)) + +#define LSXL_POL_VAL_LOW (_BIT(GPIO_FAN_HIGH) \ + | _BIT(GPIO_FAN_LOW)) + +#define LSXL_POL_VAL_HIGH (_BIT(GPIO_FUNC_LED - 32) \ + | _BIT(GPIO_ALARM_LED - 32) \ + | _BIT(GPIO_INFO_LED - 32) \ + | _BIT(GPIO_POWER_LED - 32) \ + | _BIT(GPIO_FUNC_BUTTON - 32) \ + | _BIT(GPIO_POWER_SWITCH - 32) \ + | _BIT(GPIO_POWER_AUTO_SWITCH - 32) \ + | _BIT(GPIO_FUNC_RED_LED - 32)) + +#endif /* __LSXL_H */ boards.cfg +0
-0@@ -137,6 +137,8 @@ hawkboard_uart arm arm926ejsda8xxevm davinci
enbw_cmcarm arm926ejsenbw_cmc
enbw davinci
calimain arm arm926ejs calimain
omicron davinci
dns325 arm arm926ejs -
d-link kirkwood
+lschlv2 arm arm926ejs lsxl
buffalo kirkwood lsxl:LSCHLV2
+lsxhl arm arm926ejs lsxl
buffalo kirkwood lsxl:LSXHLkm_kirkwood arm arm926ejs km_arm keymile kirkwood km_kirkwood:KM_DISABLE_PCI km_kirkwood_pci arm arm926ejs km_arm keymile kirkwood km_kirkwood:KM_RECONFIG_XLX mgcoge3un arm arm926ejs km_armkeymile kirkwood
include/configs/lsxl.h new file +32
-0@@ -0,0 +1,182 @@ +/* +* Copyright (c) 2012 Michael Walle +* Michael Walle <[email protected]> +* +* See file CREDITS for list of people who contributed to this +* project. +* +* This program is free software; you can redistribute it and/or +* modify it under the terms of the GNU General Public License as +* published by the Free Software Foundation; either version 2 of +* the License, or (at your option) any later version. +* +* This program is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +* GNU General Public License for more details. +* +* You should have received a copy of the GNU General Public License +* along with this program; if not, write to the Free Software +* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, +* MA 02110-1301 USA +*/ + +#ifndef _CONFIG_LSXL_H +#define _CONFIG_LSXL_H + +/* +* Version number information +*/ +#if defined(CONFIG_LSCHLV2 ) +#define CONFIG_IDENT_STRING " LS-CHLv2" +#define CONFIG_SYS_KWD_CONFIG $(SRCTREE)/$(CONFIG_BOARDDIR)/kwbimage-lschl.cfg
include/configs/lsxl.h new file +5
-0+#define CONFIG_MACH_TYPE 3006 +#define CONFIG_SYS_TCLK 166666667/* 166 MHz */ +#elif defined(CONFIG_LSXHL ) +#define CONFIG_IDENT_STRING " LS-XHL" +#define CONFIG_SYS_KWD_CONFIG $(SRCTREE)/$(CONFIG_BOARDDIR)/kwbimage-lsxhl.cfg
+#defineCONFIG_MACH_TYPE2663
PATCH v5 5/5 Kirkwood: add lschlv2 and lsxhl board support
Hi Prafulla,
thanks for the review.
Quoted 21 lines · Prafulla Wadaskar
Quoted 13 lines
-----Original Message-----
From: Michael Walle [mailto:michael at walle.cc]
Sent: 12 May 2012 04:21
To: u-boot at lists.denx.de
Cc: Prafulla Wadaskar; Wolfgang Denk; Mike Frysinger; Joe Hershberger; Michael Walle
Subject: [PATCH v5 5/5] Kirkwood: add lschlv2 and lsxhl board supportThis patch adds support for both the Linkstation Live (LS-CHLv2) and Linkstation Pro (LS-XHL) by Buffalo.
Signed-off-by Michael Walle <[email protected]>Cc Prafulla Wadaskar <[email protected]>Quoted 7 lines
--- MAINTAINERS | 5 + board/buffalo/lsxl/Makefile | 50 ++++++ board/buffalo/lsxl/kwbimage-lschl.cfg | 229 +++++++++++++++++++++++++ board/buffalo/lsxl/kwbimage-lsxhl.cfg | 229+++++++++++++++++++++++++
Quoted 1 line
BTW: These two file looks similar, what is the difference?
Mainly the memory configuration, eg LSCHLv2 has one 16x 512MBit, whereas the LSXHL has two 8x 1GBit dram chips. And of course some memory parameters are different.
[..snip..]
Quoted 9 lines
Quoted 9 lines
board/buffalo/lsxl/lsxl.c new file +9
-0+return 0 ; +} + +#ifdef CONFIG_MISC_INIT_R +void check_enetaddr (void ) +{ +uchar enetaddr[6 ]; + +if (!eth_getenv_enetaddr_by_index ("eth" ,0 , enetaddr)) {
Quoted 8 lines
board/buffalo/lsxl/lsxl.c line 205
Quoted 6 lines
board/buffalo/lsxl/lsxl.c new file +6
-0+/* signal unset/invalid ethaddr to user */ +set_led (LED_INFO_BLINKING ); + } +} + +static void erase_environment (void )Why do you need this function?
Mh? To erase the environment; I don't think that is the answer you wanted to know ;) Remember the only recovery method for a totally messed up environment is:
- push button longer than 10s -> erases the environment, resets the board
- push button longer than 3s -> boots into rescue mode, eg. set temporary
enetaddr and enable netconsole.
So i can't rely on a functional environment, that is i can't run stored commands. Another way would be to run a command directly, eg.
run("sf probe 0 && sf erase 70000 +10000 && reset")
dunno if that is better, back then i felt like hardcoding is better. Esp. since there may be some side effects if some special environment variables are set. Eg. there may be the spi speed set in the environment. I don't know if this is true now, but i guess one may come up with such a feature. Then this might break my recovery procedure.
Quoted 51 lines
Quoted 51 lines
board/buffalo/lsxl/lsxl.c new file +51
-0+{ +struct spi_flash *flash; + +printf ("Erasing environment..\n" ); + flash =spi_flash_probe (0 ,0 ,1000000 ,SPI_MODE_3 ); +if (!flash) { +printf ("Erasing flash failed\n" ); +return ; + } + +spi_flash_erase (flash,CONFIG_ENV_OFFSET ,CONFIG_ENV_SIZE ); +spi_flash_free (flash); +do_reset (NULL ,0 ,0 ,NULL ); +} + +static void rescue_mode (void ) +{ +uchar enetaddr[6 ]; + +printf ("Entering rescue mode..\n" ); +if (!eth_getenv_enetaddr_by_index ("eth" ,0 , enetaddr)) { +eth_random_enetaddr (enetaddr); +if (eth_setenv_enetaddr_by_index ("eth" ,0 , enetaddr)) { +printf ("Failed to set ethernet address\n" ); +set_led (LED_ALARM_BLINKING ); +return ; + } + } +setenv ("bootsource" ,"rescue" ); +} + +static void check_push_button (void ) +{ +int i =0 ; + +while (!kw_gpio_get_value (GPIO_FUNC_BUTTON )) { +udelay (100000 ); + i++; + +if (i ==10 ) +set_led (LED_INFO_ON ); + +if (i >=100 ) { +set_led (LED_INFO_BLINKING ); +break ; + } + } + +if (i >=100 ) +erase_environment (); +else if (i >=10 )
Quoted 8 lines
Review
Quoted 2 lines
(quoted diff) +2
-0+ rescue_mode(); +}You can overload this functionality to Reset key, you don't need any supporting code for the same, you can handle it through additional environment variables.
For more info check for "SYSTSTn Duration counter support" inKirkwood/cpu.c
I saw that code, but unfortunately this is not the reset button.
Quoted 5 lines
board/buffalo/lsxl/lsxl.c line 262
Quoted 4 lines
board/buffalo/lsxl/lsxl.c new file +4
-0+ +int misc_init_r (void ) +{ +check_enetaddr ();This need to be ifdefed with
CONFIG_MISC_INIT_R
it is, see above. check_enetaddr() and check_push_button() are guarded, too.
Quoted 6 lines
Quoted 6 lines
board/buffalo/lsxl/lsxl.c new file +6
-0+check_push_button (); + + return 0; +} +#endif +
[..snip..]
Quoted 8 lines
Quoted 8 lines
include/configs/lsxl.h new file +8
-0+/* +* Environment variables configurations +*/ +#ifdef CONFIG_SPI_FLASH +#define CONFIG_SYS_MAX_FLASH_BANKS 1 +#define CONFIG_SYS_MAX_FLASH_SECT 8 +#define CONFIG_ENV_IS_IN_SPI_FLASH 1 +#define CONFIG_ENV_SECT_SIZE 0x10000/* 64K */
Quoted 8 lines
include/configs/lsxl.h line 121
Quoted 6 lines
include/configs/lsxl.h new file +6
-0+#else +#define CONFIG_ENV_IS_NOWHERE +#endif + +#define CONFIG_ENV_SIZE 0x10000/* 64k */ +#define CONFIG_ENV_OFFSET 0x70000/* env starts here */ Why this is too far, what it u-boot.bin size?
Thats the original flash layout, 0x0-0x6ffff is uboot and 0x70000-0x7ffff is environment. ATM uboot size is around 40000 (hex) bytes.
Signature · 1 line
michael
PATCH v5 5/5 Kirkwood: add lschlv2 and lsxhl board support
In reply to Prafulla WadaskarAm Donnerstag 24 Mai 2012, 09:24:40 schrieb Prafulla Wadaskar:
Quoted 8 lines
Quoted 8 lines
include/configs/lsxl.h new file +6
-0+/* +* Version number information +*/ +#if defined(CONFIG_LSCHLV2 ) +#define CONFIG_IDENT_STRING " LS-CHLv2" +#define CONFIG_SYS_KWD_CONFIG $(SRCTREE)/$(CONFIG_BOARDDIR)/kwbimage-lschl.cfg
+#defineCONFIG_MACH_TYPE3006
Quoted 8 lines
Quoted 6 lines
include/configs/lsxl.h new file +4
-0+#define CONFIG_SYS_TCLK 166666667/* 166 MHz */ +#elif defined(CONFIG_LSXHL ) +#define CONFIG_IDENT_STRING " LS-XHL" +#define CONFIG_SYS_KWD_CONFIG $(SRCTREE)/$(CONFIG_BOARDDIR)/kwbimage-lsxhl.cfg
+#defineCONFIG_MACH_TYPE2663From where these number are coming for
MACH_TYPE?
Sorry forgot to answer that. They come from here:
http://www.arm.linux.org.uk/developer/machines/download.php
Signature · 1 line
Michael
v5 cover Kirkwood: add lschlv2 and lsxhl board support
In reply to Michael WalleQuoted 8 lines
-----Original Message-----
From: Michael Walle [mailto:michael at walle.cc]
Sent: 12 May 2012 04:21
To: u-boot at lists.denx.de
Cc: Prafulla Wadaskar; Wolfgang Denk; Mike Frysinger; Joe Hershberger; Michael Walle
Subject: [PATCH v5 0/5] Kirkwood: add lschlv2 and lsxhl board support
Quoted 4 lines
Changes:
v5:
- combine patchset again. the "net: *" patches should be individually acked by net custodian
Hi Joe,
I have not issues pulling this patch series but most of the patches are going for net related changes.
I would like you have your final ack on these.
Regards..
Prafulla . . .
v5 cover Kirkwood: add lschlv2 and lsxhl board support
Hi Prafulla,
Quoted 13 lines · Prafulla Wadaskar
Quoted 8 lines
-----Original Message-----
From: Michael Walle [mailto:michael at walle.cc]
Sent: 12 May 2012 04:21
To: u-boot at lists.denx.de
Cc: Prafulla Wadaskar; Wolfgang Denk; Mike Frysinger; Joe Hershberger; Michael Walle
Subject: [PATCH v5 0/5] Kirkwood: add lschlv2 and lsxhl board supportQuoted 4 lines
Changes:
v5:
?- combine patchset again. the "net: *" patches should be individually ? ?acked by net custodian
Quoted 4 lines
Hi Joe,
I have not issues pulling this patch series but most of the patches are going for net related changes.I would like you have your final ack on these.
OK. I'll have a look again.
-Joe