imx: Align the imximage header and payload to multiples of 4k
PATCH 1/6 imx: Align the imximage header and payload to multiples of 4k
The MX53 ROM loads the data from NAND in multiples of pages and supports maximum page size of 4k. Thus, align the image and header to 4k to be safe from ROM bugs.
--- tools/imximage.c | 11 +++++++---- tools/imximage.h | 3 ++- 2 files changed, 9 insertions(+), 5 deletions(-)
tools/imximage.c +7 -4
@@ -518,11 +518,14 @@ static void imximage_set_header(void *ptr, struct stat *sbuf, int ifd,/* * ROM bug alert -* mx53 only loads 512 byte multiples. -* The remaining fraction of a block bytes would -* not be loaded. +* +* MX53 only loads 512 byte multiples in case of SD boot. +* MX53 only loads NAND page multiples in case of NAND boot and +* supports up to 4096 byte large pages, thus align to 4096. +* +* The remaining fraction of a block bytes would not be loaded! */ - *header_size_ptr =ROUND (sbuf->st_size + imxhdr->flash_offset,512 ); + *header_size_ptr =ROUND (sbuf->st_size + imxhdr->flash_offset,4096 ); }int imximage_check_params (struct mkimage_params *params)
tools/imximage.h +2 -1
@@ -151,13 +151,14 @@ typedef struct {dcd_v2_t dcd_table; } imx_header_v2_t; +/* The header must be aligned to 4k on MX53 for NAND boot */ struct imx_header {union {imx_header_v1_t hdr_v1;imx_header_v2_t hdr_v2; } header;uint32_t flash_offset; -}; +} __attribute__((aligned (4096 ))); typedef void (*set_dcd_val_t)(struct imx_header *imxhdr, char *name, int lineno,
Signature · 1 line
1.7.11.7
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
Add support for generic NAND SPL via the SPL framework into the mxc_nand_spl driver. This is basically just a simple rename and publication of the already implemented functions. To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macro CONFIG_SPL_NAND_LEGACY was introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.
Also make sure the requested payload is aligned to full pages, otherwise this simple driver fails to load the last page.
--- drivers/mtd/nand/mxc_nand_spl.c | 13 ++++++++++--- include/configs/mx31pdk.h | 1 + include/configs/tx25.h | 1 + 3 files changed, 12 insertions(+), 3 deletions(-)
drivers/mtd/nand/mxc_nand_spl.c +10 -3
@@ -290,7 +290,7 @@ static int is_badblock(int pagenumber)return 0 ; } -static int nand_load (unsigned int from,unsigned int size,unsigned char *buf) +int nand_spl_load_image (uint32_t from,unsigned int size,void *buf) {int i;unsigned int page;
@@ -303,6 +303,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) page = from /CONFIG_SYS_NAND_PAGE_SIZE ; i =0 ; + size =roundup (size,CONFIG_SYS_NAND_PAGE_SIZE );while (i < size /CONFIG_SYS_NAND_PAGE_SIZE ) {if (nfc_read_page (page, buf) <0 )return -1 ;
@@ -332,6 +333,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf)return 0 ; } +#ifdef CONFIG_SPL_NAND_SUPPORT_LEGACY /* * The main entry for NAND booting. It's necessary that SDRAM is already * configured and available since this code loads the main U-Boot image
@@ -345,8 +347,9 @@ void nand_boot(void)* CONFIG_SYS_NAND_U_BOOT_OFFS and CONFIG_SYS_NAND_U_BOOT_SIZE must * be aligned to full pages */ -if (!nand_load (CONFIG_SYS_NAND_U_BOOT_OFFS ,CONFIG_SYS_NAND_U_BOOT_SIZE , - (uchar *)CONFIG_SYS_NAND_U_BOOT_DST )) { +if (!nand_spl_load_image (CONFIG_SYS_NAND_U_BOOT_OFFS , +CONFIG_SYS_NAND_U_BOOT_SIZE , + (uchar *)CONFIG_SYS_NAND_U_BOOT_DST )) {/* Copy from NAND successful, start U-boot */ uboot = (void *)CONFIG_SYS_NAND_U_BOOT_START ;uboot ();
@@ -364,3 +367,7 @@ void hang(void)/* Loop forever */ while (1) ; } +#endif + +void nand_init (void ) {} +void nand_deselect (void ) {}
include/configs/mx31pdk.h +1 -0
@@ -50,6 +50,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-spl.lds"#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define CONFIG_SPL_TEXT_BASE 0x87dc0000#define CONFIG_SYS_TEXT_BASE 0x87e00000
include/configs/tx25.h +1 -0
@@ -37,6 +37,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-spl.lds"#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define CONFIG_SPL_TEXT_BASE 0x810c0000#define CONFIG_SYS_TEXT_BASE 0x81200000
Signature · 1 line
1.7.11.7
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
Dear Marek Vasut,
Quoted 103 lines · Marek Vasut
Add support for generic NAND SPL via the SPL framework into the
mxc_nand_spldriver. This is basically just a simple rename and publication of the already implemented functions. To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macroCONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.Also make sure the requested payload is aligned to full pages, otherwise this simple driver fails to load the last page.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>--- drivers/mtd/nand/mxc_nand_spl.c | 13 ++++++++++--- include/configs/mx31pdk.h | 1 + include/configs/tx25.h | 1 + 3 files changed, 12 insertions(+), 3 deletions(-)drivers/mtd/nand/mxc_nand_spl.c +1
-1@@ -290,7 +290,7 @@ static int is_badblock(int pagenumber)return 0 ; } -static int nand_load (unsigned int from,unsigned int size,unsigned char *buf) +int nand_spl_load_image (uint32_t from,unsigned int size,void *buf) {int i;drivers/mtd/nand/mxc_nand_spl.c +1
-0unsigned int page;@@ -303,6 +303,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) page = from /CONFIG_SYS_NAND_PAGE_SIZE ; i =0 ; + size =roundup (size,CONFIG_SYS_NAND_PAGE_SIZE );while (i < size /CONFIG_SYS_NAND_PAGE_SIZE ) {if (nfc_read_page (page, buf) <0 )drivers/mtd/nand/mxc_nand_spl.c +1
-0return -1 ;@@ -332,6 +333,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf)return 0 ; } +#ifdef CONFIG_SPL_NAND_SUPPORT_LEGACY /* * The main entry for NAND booting. It's necessary that SDRAM is already * configured and available since this code loads the main U-Boot image
drivers/mtd/nand/mxc_nand_spl.c +7
-2@@ -345,8 +347,9 @@ void nand_boot(void)* CONFIG_SYS_NAND_U_BOOT_OFFS and CONFIG_SYS_NAND_U_BOOT_SIZE must * be aligned to full pages */ -if (!nand_load (CONFIG_SYS_NAND_U_BOOT_OFFS ,CONFIG_SYS_NAND_U_BOOT_SIZE , - (uchar *)CONFIG_SYS_NAND_U_BOOT_DST )) { +if (!nand_spl_load_image (CONFIG_SYS_NAND_U_BOOT_OFFS , +CONFIG_SYS_NAND_U_BOOT_SIZE , + (uchar *)CONFIG_SYS_NAND_U_BOOT_DST )) {/* Copy from NAND successful, start U-boot */ uboot = (void *)CONFIG_SYS_NAND_U_BOOT_START ;uboot ();@@ -364,3 +367,7 @@ void hang(void)/* Loop forever */ while (1) ; } +#endif + +void nand_init (void ) {} +void nand_deselect (void ) {}include/configs/mx31pdk.h +1
-0@@ -50,6 +50,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-spl.lds"#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define CONFIG_SPL_TEXT_BASE 0x87dc0000#define CONFIG_SYS_TEXT_BASE 0x87e00000include/configs/tx25.h +1
-1@@ -37,6 +37,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-spl.lds"#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define CONFIG_SPL_TEXT_BASE 0x810c0000#define CONFIG_SYS_TEXT_BASE 0x81200000 --
Quoted 1 line
1.7.11.7
This is not about legacy vs. non-legacy. This is about basic vs. more featured SPL because of SPL size constraints. So what about dropping CONFIG_SPL_NAND_SUPPORT_LEGACY and testing for CONFIG_SPL_FRAMEWORK definition instead?
Best regards,
Beno?t
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
Quoted 103 lines · Beno?t Th?baudeau
Dear Marek Vasut,
Quoted 100 lines · Marek Vasut
Add support for generic NAND SPL via the SPL framework into the
mxc_nand_spldriver. This is basically just a simple rename and publication of the already implemented functions. To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macroCONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.Also make sure the requested payload is aligned to full pages, otherwise this simple driver fails to load the last page.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>--- drivers/mtd/nand/mxc_nand_spl.c | 13 ++++++++++--- include/configs/mx31pdk.h | 1 + include/configs/tx25.h | 1 + 3 files changed, 12 insertions(+), 3 deletions(-)drivers/mtd/nand/mxc_nand_spl.c +1
-1@@ -290,7 +290,7 @@ static int is_badblock(int pagenumber)return 0 ; } -static int nand_load (unsigned int from,unsigned int size,unsigned char *buf) +int nand_spl_load_image (uint32_t from,unsigned int size,void *buf) {int i;drivers/mtd/nand/mxc_nand_spl.c +1
-0unsigned int page;@@ -303,6 +303,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) page = from /CONFIG_SYS_NAND_PAGE_SIZE ; i =0 ; + size =roundup (size,CONFIG_SYS_NAND_PAGE_SIZE );while (i < size /CONFIG_SYS_NAND_PAGE_SIZE ) {if (nfc_read_page (page, buf) <0 )drivers/mtd/nand/mxc_nand_spl.c +1
-0return -1 ;@@ -332,6 +333,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf)return 0 ; } +#ifdef CONFIG_SPL_NAND_SUPPORT_LEGACY /* * The main entry for NAND booting. It's necessary that SDRAM is already * configured and available since this code loads the main U-Boot image
drivers/mtd/nand/mxc_nand_spl.c +7
-2@@ -345,8 +347,9 @@ void nand_boot(void)* CONFIG_SYS_NAND_U_BOOT_OFFS and CONFIG_SYS_NAND_U_BOOT_SIZE must * be aligned to full pages */ -if (!nand_load (CONFIG_SYS_NAND_U_BOOT_OFFS ,CONFIG_SYS_NAND_U_BOOT_SIZE , - (uchar *)CONFIG_SYS_NAND_U_BOOT_DST )) { +if (!nand_spl_load_image (CONFIG_SYS_NAND_U_BOOT_OFFS , +CONFIG_SYS_NAND_U_BOOT_SIZE , + (uchar *)CONFIG_SYS_NAND_U_BOOT_DST )) {/* Copy from NAND successful, start U-boot */ uboot = (void *)CONFIG_SYS_NAND_U_BOOT_START ;uboot ();@@ -364,3 +367,7 @@ void hang(void)/* Loop forever */ while (1) ; } +#endif + +void nand_init (void ) {} +void nand_deselect (void ) {}include/configs/mx31pdk.h +1
-0@@ -50,6 +50,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-spl.lds"#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define CONFIG_SPL_TEXT_BASE 0x87dc0000#define CONFIG_SYS_TEXT_BASE 0x87e00000include/configs/tx25.h +1
-0@@ -37,6 +37,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-spl.lds"#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY
Quoted 11 lines
Quoted 4 lines
include/configs/tx25.h +0
-1#define CONFIG_SPL_TEXT_BASE 0x810c0000#define CONFIG_SYS_TEXT_BASE 0x81200000 --1.7.11.7
This is not about legacy vs. non-legacy. This is about basic vs. more featured
SPL because of SPL size constraints. So what about droppingCONFIG_SPL_NAND_SUPPORT_LEGACYand testing forCONFIG_SPL_FRAMEWORKdefinition
instead?
Also, I don't see any nand_init() (called by spl_nand_load_image()) for m53evk.
Best regards,
Beno?t
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
Dear Beno?t Th?baudeau,
Quoted 122 lines · Beno?t Th?baudeau
Quoted 121 lines
Dear Marek Vasut,
Quoted 118 lines · Marek Vasut
Add support for generic NAND SPL via the SPL framework into the
mxc_nand_spldriver. This is basically just a simple rename and publication of the already implemented functions. To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macroCONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.Also make sure the requested payload is aligned to full pages, otherwise this simple driver fails to load the last page.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>---
drivers/mtd/nand/mxc_nand_spl.c | 13 ++++++++++--- include/configs/mx31pdk.h | 1 + include/configs/tx25.h | 1 + 3 files changed, 12 insertions(+), 3 deletions(-)drivers/mtd/nand/mxc_nand_spl.c +1
-1@@ -290,7 +290,7 @@ static int is_badblock(int pagenumber)return 0 ; } -static int nand_load (unsigned int from,unsigned int size,unsigned char *buf) +int nand_spl_load_image (uint32_t from,unsigned int size,void *buf){
drivers/mtd/nand/mxc_nand_spl.c +1
-0int i;unsigned int page;@@ -303,6 +303,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) page = from /CONFIG_SYS_NAND_PAGE_SIZE ; i =0 ; + size =roundup (size,CONFIG_SYS_NAND_PAGE_SIZE );drivers/mtd/nand/mxc_nand_spl.c +1
-0while (i < size /CONFIG_SYS_NAND_PAGE_SIZE ) {if (nfc_read_page (page, buf) <0 )return -1 ;@@ -332,6 +333,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) return 0; } +#ifdef CONFIG_SPL_NAND_SUPPORT_LEGACY /*
* The main entry for NAND booting. It's necessary that SDRAM is already * configured and available since this code loads the main U-Boot image
drivers/mtd/nand/mxc_nand_spl.c +0
-1@@ -345,8 +347,9 @@ void nand_boot(void)* CONFIG_SYS_NAND_U_BOOT_OFFS and CONFIG_SYS_NAND_U_BOOT_SIZE must * be aligned to full pages */ - if (!nand_load(CONFIG_SYS_NAND_U_BOOT_OFFS,
CONFIG_SYS_NAND_U_BOOT_SIZE, - (uchar
*)CONFIG_SYS_NAND_U_BOOT_DST)) {drivers/mtd/nand/mxc_nand_spl.c +7
-0+if (!nand_spl_load_image (CONFIG_SYS_NAND_U_BOOT_OFFS , +CONFIG_SYS_NAND_U_BOOT_SIZE , + (uchar *)CONFIG_SYS_NAND_U_BOOT_DST )) {/* Copy from NAND successful, start U-boot */ uboot = (void *)CONFIG_SYS_NAND_U_BOOT_START ;uboot ();@@ -364,3 +367,7 @@ void hang(void)/* Loop forever */ while (1) ; } +#endif + +void nand_init (void ) {} +void nand_deselect (void ) {}include/configs/mx31pdk.h +1
-0@@ -50,6 +50,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-spl.lds"#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x87dc0000
#defineCONFIG_SYS_TEXT_BASE0x87e00000
Quoted 7 lines
Quoted 7 lines
include/configs/tx25.h line 38
Quoted 7 lines
include/configs/tx25.h +0
-0@@ -37,6 +37,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-
spl.lds"
Quoted 11 lines
Quoted 11 lines
Quoted 10 lines
include/configs/tx25.h +1
-0#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x810c0000
#defineCONFIG_SYS_TEXT_BASE0x81200000--
1.7.11.7
Quoted 6 lines
Quoted 6 lines
This is not about legacy vs. non-legacy. This is about basic vs. more featured
SPL because of SPL size constraints. So what about droppingCONFIG_SPL_NAND_SUPPORT_LEGACYand testing forCONFIG_SPL_FRAMEWORKdefinition
instead?
I was thinking about that, but the symbol is unrelated to NAND. I still think it's either a matter of fixing for new SPL or removing those two boards. The nand_spl/ stuff shall be removed ASAP.
Quoted 2 lines
Also, I don't see any
nand_init()(called byspl_nand_load_image()) for m53evk.
What do you mean?
Best regards,
Marek Vasut
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
Dear Marek Vasut,
Quoted 143 lines · Marek Vasut
Dear Beno?t Th?baudeau,
Quoted 122 lines · Beno?t Th?baudeau
Quoted 121 lines
Dear Marek Vasut,
Quoted 118 lines · Marek Vasut
Add support for generic NAND SPL via the SPL framework into the
mxc_nand_spldriver. This is basically just a simple rename and publication of the already implemented functions. To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macroCONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.Also make sure the requested payload is aligned to full pages, otherwise this simple driver fails to load the last page.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>---
drivers/mtd/nand/mxc_nand_spl.c | 13 ++++++++++--- include/configs/mx31pdk.h | 1 + include/configs/tx25.h | 1 + 3 files changed, 12 insertions(+), 3 deletions(-)drivers/mtd/nand/mxc_nand_spl.c +1
-1@@ -290,7 +290,7 @@ static int is_badblock(int pagenumber)return 0 ; } -static int nand_load (unsigned int from,unsigned int size,unsigned char *buf) +int nand_spl_load_image (uint32_t from,unsigned int size,void *buf){
drivers/mtd/nand/mxc_nand_spl.c +1
-0int i;unsigned int page;@@ -303,6 +303,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) page = from /CONFIG_SYS_NAND_PAGE_SIZE ; i =0 ; + size =roundup (size,CONFIG_SYS_NAND_PAGE_SIZE );drivers/mtd/nand/mxc_nand_spl.c +1
-0while (i < size /CONFIG_SYS_NAND_PAGE_SIZE ) {if (nfc_read_page (page, buf) <0 )return -1 ;@@ -332,6 +333,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) return 0; } +#ifdef CONFIG_SPL_NAND_SUPPORT_LEGACY /*
* The main entry for NAND booting. It's necessary that SDRAM is already * configured and available since this code loads the main U-Boot image
drivers/mtd/nand/mxc_nand_spl.c +0
-1@@ -345,8 +347,9 @@ void nand_boot(void)* CONFIG_SYS_NAND_U_BOOT_OFFS and CONFIG_SYS_NAND_U_BOOT_SIZE must * be aligned to full pages */ - if (!nand_load(CONFIG_SYS_NAND_U_BOOT_OFFS,
CONFIG_SYS_NAND_U_BOOT_SIZE, - (uchar
*)CONFIG_SYS_NAND_U_BOOT_DST)) {drivers/mtd/nand/mxc_nand_spl.c +7
-0+if (!nand_spl_load_image (CONFIG_SYS_NAND_U_BOOT_OFFS , +CONFIG_SYS_NAND_U_BOOT_SIZE , + (uchar *)CONFIG_SYS_NAND_U_BOOT_DST )) {/* Copy from NAND successful, start U-boot */ uboot = (void *)CONFIG_SYS_NAND_U_BOOT_START ;uboot ();@@ -364,3 +367,7 @@ void hang(void)/* Loop forever */ while (1) ; } +#endif + +void nand_init (void ) {} +void nand_deselect (void ) {}include/configs/mx31pdk.h +1
-0@@ -50,6 +50,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-spl.lds"#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x87dc0000
#defineCONFIG_SYS_TEXT_BASE0x87e00000Quoted 7 lines
Quoted 7 lines
include/configs/tx25.h line 38
Quoted 7 lines
include/configs/tx25.h +0
-0@@ -37,6 +37,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-
spl.lds"Quoted 11 lines
Quoted 11 lines
Quoted 10 lines
include/configs/tx25.h +1
-0#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x810c0000
#defineCONFIG_SYS_TEXT_BASE0x81200000--
1.7.11.7
Quoted 8 lines
Quoted 6 lines
Quoted 6 lines
This is not about legacy vs. non-legacy. This is about basic vs. more featured
SPL because of SPL size constraints. So what about droppingCONFIG_SPL_NAND_SUPPORT_LEGACYand testing forCONFIG_SPL_FRAMEWORKdefinition
instead?I was thinking about that, but the symbol is unrelated to NAND.
Well, it's CONFIG_SPL_FRAMEWORK + CONFIG_SPL_NAND_SUPPORT that defines the other implementation, and CONFIG_SPL_NAND_SUPPORT triggers the build of mxc_nand_spl.c for SPL, so the common point is CONFIG_SPL_FRAMEWORK.
Quoted 3 lines
I still think
it's either a matter of fixing for new SPL or removing those two boards. The nand_spl/ stuff shall be removed ASAP.
Removing those boards is not a solution.
Is it really about "new" SPL? The generic SPL is enabled by CONFIG_SPL, and CONFIG_SPL_FRAMEWORK is a sub-option. If the generic SPL rules imposed to use the SPL framework functions, there would be no such sub-option. So it looks like these boards are complying to the new SPL rules.
We could see if using CONFIG_SPL_FRAMEWORK would still allow the SPL to fit in 2 kiB in order to drop this function, but if it does not fit, the new SPL rules should still make it possible to have a solution for any board having SPL size constraints.
Quoted 4 lines
Quoted 2 lines
Also, I don't see any
nand_init()(called byspl_nand_load_image()) for m53evk.What do you mean?
I mean that I don't identify the implementation of nand_init() used in the context of the m53evk since the builds of drivers/mtd/nand/nand.c and drivers/mtd/nand/nand_spl_simple.c do not seem to be enabled for this board, and I don't see another definition.
Best regards,
Beno?t
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
Dear Marek Vasut,
Quoted 174 lines · Beno?t Th?baudeau
Dear Marek Vasut,
Quoted 143 lines · Marek Vasut
Dear Beno?t Th?baudeau,
Quoted 122 lines · Beno?t Th?baudeau
Quoted 121 lines
Dear Marek Vasut,
Quoted 118 lines · Marek Vasut
Add support for generic NAND SPL via the SPL framework into the
mxc_nand_spldriver. This is basically just a simple rename and publication of the already implemented functions. To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macroCONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.Also make sure the requested payload is aligned to full pages, otherwise this simple driver fails to load the last page.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>---
drivers/mtd/nand/mxc_nand_spl.c | 13 ++++++++++--- include/configs/mx31pdk.h | 1 + include/configs/tx25.h | 1 + 3 files changed, 12 insertions(+), 3 deletions(-)drivers/mtd/nand/mxc_nand_spl.c +1
-1@@ -290,7 +290,7 @@ static int is_badblock(int pagenumber)return 0 ; } -static int nand_load (unsigned int from,unsigned int size,unsigned char *buf) +int nand_spl_load_image (uint32_t from,unsigned int size,void *buf){
drivers/mtd/nand/mxc_nand_spl.c +1
-0int i;unsigned int page;@@ -303,6 +303,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) page = from /CONFIG_SYS_NAND_PAGE_SIZE ; i =0 ; + size =roundup (size,CONFIG_SYS_NAND_PAGE_SIZE );drivers/mtd/nand/mxc_nand_spl.c +1
-0while (i < size /CONFIG_SYS_NAND_PAGE_SIZE ) {if (nfc_read_page (page, buf) <0 )return -1 ;@@ -332,6 +333,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) return 0; } +#ifdef CONFIG_SPL_NAND_SUPPORT_LEGACY /*
* The main entry for NAND booting. It's necessary that SDRAM is already * configured and available since this code loads the main U-Boot image
drivers/mtd/nand/mxc_nand_spl.c +0
-1@@ -345,8 +347,9 @@ void nand_boot(void)* CONFIG_SYS_NAND_U_BOOT_OFFS and CONFIG_SYS_NAND_U_BOOT_SIZE must * be aligned to full pages */ - if (!nand_load(CONFIG_SYS_NAND_U_BOOT_OFFS,
CONFIG_SYS_NAND_U_BOOT_SIZE, - (uchar
*)CONFIG_SYS_NAND_U_BOOT_DST)) {drivers/mtd/nand/mxc_nand_spl.c +7
-0+if (!nand_spl_load_image (CONFIG_SYS_NAND_U_BOOT_OFFS , +CONFIG_SYS_NAND_U_BOOT_SIZE , + (uchar *)CONFIG_SYS_NAND_U_BOOT_DST )) {/* Copy from NAND successful, start U-boot */ uboot = (void *)CONFIG_SYS_NAND_U_BOOT_START ;uboot ();@@ -364,3 +367,7 @@ void hang(void)/* Loop forever */ while (1) ; } +#endif + +void nand_init (void ) {} +void nand_deselect (void ) {}include/configs/mx31pdk.h +1
-0@@ -50,6 +50,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-spl.lds"#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x87dc0000
#defineCONFIG_SYS_TEXT_BASE0x87e00000Quoted 7 lines
Quoted 7 lines
include/configs/tx25.h line 38
Quoted 7 lines
include/configs/tx25.h +0
-0@@ -37,6 +37,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-
spl.lds"Quoted 11 lines
Quoted 11 lines
Quoted 10 lines
include/configs/tx25.h +1
-0#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x810c0000
#defineCONFIG_SYS_TEXT_BASE0x81200000--
1.7.11.7Quoted 8 lines
Quoted 6 lines
Quoted 6 lines
This is not about legacy vs. non-legacy. This is about basic vs. more featured
SPL because of SPL size constraints. So what about droppingCONFIG_SPL_NAND_SUPPORT_LEGACYand testing forCONFIG_SPL_FRAMEWORKdefinition
instead?I was thinking about that, but the symbol is unrelated to NAND.
Well, it's
CONFIG_SPL_FRAMEWORK+CONFIG_SPL_NAND_SUPPORTthat defines the other
implementation, andCONFIG_SPL_NAND_SUPPORTtriggers the build ofmxc_nand_spl.c
for SPL, so the common point isCONFIG_SPL_FRAMEWORK.Quoted 4 lines
I still think
it's either a matter of fixing for new SPL or removing those two boards. The
nand_spl/ stuff shall be removed ASAP.Removing those boards is not a solution.
Is it really about "new" SPL? The generic SPL is enabled by
CONFIG_SPL, andCONFIG_SPL_FRAMEWORKis a sub-option. If the generic SPL rules imposed to use the SPL framework functions, there would be no such sub-option. So it looks like
these boards are complying to the new SPL rules.
Quoted 7 lines
We could see if using
CONFIG_SPL_FRAMEWORKwould still allow the SPL to fit in
2 kiB in order to drop this function, but if it does not fit, the new SPL rules
should still make it possible to have a solution for any board having SPL size
constraints.
I've just quickly build-tested the hack-patch below to have a guess of the SPL size for mx31pdk with CONFIG_SPL_FRAMEWORK, and I get 3.9 kiB, so way more than the maximum 2 kiB. Hence, these boards can't unfortunately be made more generic, but it's not a reason to remove them. The generic SPL must allow such hardware to be supported by U-Boot.
Quoted 10 lines
Quoted 4 lines
Quoted 2 lines
Also, I don't see any
nand_init()(called byspl_nand_load_image()) for m53evk.What do you mean?
I mean that I don't identify the implementation of
nand_init()used in the context of the m53evk since the builds ofdrivers/mtd/nand/nand.canddrivers/mtd/nand/nand_spl_simple.cdo not seem to be enabled for this board, and
I don't see another definition.
Forget that. I had just missed its definition in this patch. I saw it thanks to Philip.
Best regards,
Beno?t
---
board/freescale/mx31pdk/mx31pdk.c +7 -1
@@ -33,6 +33,7 @@#include <power/pmic.h> #include <fsl_pmic.h> #include <errno.h> +#include <spl.h> DECLARE_GLOBAL_DATA_PTR ;
@@ -40,7 +41,12 @@ DECLARE_GLOBAL_DATA_PTR;void board_init_f (ulong bootflag) {relocate_code (CONFIG_SPL_TEXT_BASE ); - asmvolatile ("ldr pc, =nand_boot" ); + asmvolatile ("ldr pc, =board_init_r" ); +} + +u32 spl_boot_device (void ) +{ +return BOOT_DEVICE_NAND ; }#endif
drivers/mtd/nand/mxc_nand_spl.c +3 -33
@@ -290,7 +290,7 @@ static int is_badblock(int pagenumber)return 0 ; } -static int nand_load (unsigned int from,unsigned int size,unsigned char *buf) +int nand_spl_load_image (uint32_t from,unsigned int size,void *buf) {int i;unsigned int page;
@@ -332,35 +332,5 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf)return 0 ; } -/* -* The main entry for NAND booting. It's necessary that SDRAM is already -* configured and available since this code loads the main U-Boot image -* from NAND into SDRAM and starts it from there. -*/ -void nand_boot (void ) -{ - __attribute__((noreturn))void (*uboot)(void ); - -/* -* CONFIG_SYS_NAND_U_BOOT_OFFS and CONFIG_SYS_NAND_U_BOOT_SIZE must -* be aligned to full pages -*/ -if (!nand_load (CONFIG_SYS_NAND_U_BOOT_OFFS ,CONFIG_SYS_NAND_U_BOOT_SIZE , - (uchar *)CONFIG_SYS_NAND_U_BOOT_DST )) { -/* Copy from NAND successful, start U-boot */ - uboot = (void *)CONFIG_SYS_NAND_U_BOOT_START ; -uboot (); - }else { -/* Unrecoverable error when copying from NAND */ -hang (); - } -} - -/* -* Called in case of an exception. -*/ -void hang (void ) -{ -/* Loop forever */ -while (1 ) ; -} +void nand_init (void ) {} +void nand_deselect (void ) {}
include/configs/mx31pdk.h +5 -0
@@ -50,6 +50,11 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-spl.lds"#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_FRAMEWORK +#define CONFIG_SPL_STACK CONFIG_SYS_INIT_SP_ADDR +#define CONFIG_SPL_LIBCOMMON_SUPPORT +#define CONFIG_SPL_LIBGENERIC_SUPPORT +#define CONFIG_SPL_SERIAL_SUPPORT #define CONFIG_SPL_TEXT_BASE 0x87dc0000#define CONFIG_SYS_TEXT_BASE 0x87e00000
spl/Makefile +4 -0
@@ -88,6 +88,10 @@ ifneq ($(CONFIG_AM33XX)$(CONFIG_OMAP34XX)$(CONFIG_OMAP44XX)$(CONFIG_OMAP54XX)$(C LIBS-y += $(CPUDIR)/omap-common/libomap-common.o endif +ifneq (,$(filter $(SOC), mx25 mx27 mx5 mx6 mx31 mx35)) +LIBS-y += arch/$(ARCH)/imx-common/libimx-common.o +endif + ifneq ($(CONFIG_TEGRA),) LIBS-y += arch/$(ARCH)/cpu/$(SOC)-common/lib$(SOC)-common.o LIBS-y += arch/$(ARCH)/cpu/tegra-common/libcputegra-common.o
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
Dear Beno?t Th?baudeau,
Quoted 122 lines
Dear Marek Vasut,
Quoted 119 lines · Beno?t Th?baudeau
Dear Marek Vasut,
Quoted 116 lines · Marek Vasut
Dear Beno?t Th?baudeau,
Quoted 114 lines · Beno?t Th?baudeau
Quoted 113 lines
Dear Marek Vasut,
Quoted 110 lines · Marek Vasut
Add support for generic NAND SPL via the SPL framework into the
mxc_nand_spldriver. This is basically just a simple rename and publication of the already implemented functions. To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macroCONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.Also make sure the requested payload is aligned to full pages, otherwise this simple driver fails to load the last page.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>---
drivers/mtd/nand/mxc_nand_spl.c | 13 ++++++++++--- include/configs/mx31pdk.h | 1 + include/configs/tx25.h | 1 + 3 files changed, 12 insertions(+), 3 deletions(-)drivers/mtd/nand/mxc_nand_spl.c +1
-1@@ -290,7 +290,7 @@ static int is_badblock(int pagenumber)return 0 ; } -static int nand_load (unsigned int from,unsigned int size, unsigned char *buf) +int nand_spl_load_image(uint32_t from, unsigned int size, void*buf)
{
drivers/mtd/nand/mxc_nand_spl.c +1
-0int i;unsigned int page;@@ -303,6 +303,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) page = from /CONFIG_SYS_NAND_PAGE_SIZE ; i =0 ; + size =roundup (size,CONFIG_SYS_NAND_PAGE_SIZE );drivers/mtd/nand/mxc_nand_spl.c +1
-0while (i < size /CONFIG_SYS_NAND_PAGE_SIZE ) {if (nfc_read_page (page, buf) <0 )return -1 ;@@ -332,6 +333,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) return 0; } +#ifdef CONFIG_SPL_NAND_SUPPORT_LEGACY /*
* The main entry for NAND booting. It's necessary that SDRAM is already * configured and available since this code loads the main U-Boot image
drivers/mtd/nand/mxc_nand_spl.c +0
-1@@ -345,8 +347,9 @@ void nand_boot(void)* CONFIG_SYS_NAND_U_BOOT_OFFS and CONFIG_SYS_NAND_U_BOOT_SIZE must * be aligned to full pages */ - if (!nand_load(CONFIG_SYS_NAND_U_BOOT_OFFS,
CONFIG_SYS_NAND_U_BOOT_SIZE, - (uchar
*)CONFIG_SYS_NAND_U_BOOT_DST)) {drivers/mtd/nand/mxc_nand_spl.c +7
-0+if (!nand_spl_load_image (CONFIG_SYS_NAND_U_BOOT_OFFS , +CONFIG_SYS_NAND_U_BOOT_SIZE , + (uchar *)CONFIG_SYS_NAND_U_BOOT_DST )) {/* Copy from NAND successful, start U-boot */ uboot = (void *)CONFIG_SYS_NAND_U_BOOT_START ;uboot ();@@ -364,3 +367,7 @@ void hang(void)/* Loop forever */ while (1) ; } +#endif + +void nand_init (void ) {} +void nand_deselect (void ) {}(quoted diff) +0
-0
b/include/configs/mx31pdk.hindex 1754595..217552e 100644include/configs/mx31pdk.h +0
-0@@ -50,6 +50,7 @@
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
include/configs/mx31pdk.h line 51
Quoted 1 line
include/configs/mx31pdk.h +0
-0#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-
spl.lds"
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
include/configs/mx31pdk.h +1
-0#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x87dc0000
#defineCONFIG_SYS_TEXT_BASE0x87e00000
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
include/configs/tx25.h line 38
Quoted 7 lines
include/configs/tx25.h +0
-0@@ -37,6 +37,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-
boot-
Quoted 50 lines
Quoted 42 lines
Quoted 14 lines
spl.lds"Quoted 11 lines
Quoted 11 lines
Quoted 10 lines
include/configs/tx25.h +1
-0#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x810c0000
#defineCONFIG_SYS_TEXT_BASE0x81200000--
1.7.11.7Quoted 8 lines
Quoted 6 lines
Quoted 6 lines
This is not about legacy vs. non-legacy. This is about basic vs. more featured
SPL because of SPL size constraints. So what about droppingCONFIG_SPL_NAND_SUPPORT_LEGACYand testing forCONFIG_SPL_FRAMEWORKdefinition
instead?I was thinking about that, but the symbol is unrelated to NAND.
Well, it's
CONFIG_SPL_FRAMEWORK+CONFIG_SPL_NAND_SUPPORTthat defines the other
implementation, andCONFIG_SPL_NAND_SUPPORTtriggers the build ofmxc_nand_spl.c
for SPL, so the common point isCONFIG_SPL_FRAMEWORK.Quoted 4 lines
I still think
it's either a matter of fixing for new SPL or removing those two boards. The
nand_spl/ stuff shall be removed ASAP.Removing those boards is not a solution.
Is it really about "new" SPL? The generic SPL is enabled by
CONFIG_SPL, andCONFIG_SPL_FRAMEWORKis a sub-option. If the generic SPL rules imposed to use the SPL framework functions, there would be no such sub-option. So it looks like
these boards are complying to the new SPL rules.Quoted 7 lines
We could see if using
CONFIG_SPL_FRAMEWORKwould still allow the SPL to fit in
2 kiB in order to drop this function, but if it does not fit, the new SPL rules
should still make it possible to have a solution for any board having SPL size
constraints.
Quoted 5 lines
I've just quickly build-tested the hack-patch below to have a guess of the SPL size for mx31pdk with
CONFIG_SPL_FRAMEWORK, and I get 3.9 kiB, so way more than the maximum 2 kiB. Hence, these boards can't unfortunately be made more generic, but it's not a reason to remove them. The generic SPL must allow such hardware to be supported by U-Boot.
See the other email for my other idea. Either we can go with that or I can focus on making use of the regular MXC NAND driver on mx5, since there's no 2KiB limit there.
[...]
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
Dear Marek Vasut,
Quoted 198 lines · Marek Vasut
Dear Beno?t Th?baudeau,
Quoted 122 lines
Dear Marek Vasut,
Quoted 119 lines · Beno?t Th?baudeau
Dear Marek Vasut,
Quoted 116 lines · Marek Vasut
Dear Beno?t Th?baudeau,
Quoted 114 lines · Beno?t Th?baudeau
Quoted 113 lines
Dear Marek Vasut,
Quoted 110 lines · Marek Vasut
Add support for generic NAND SPL via the SPL framework into the
mxc_nand_spldriver. This is basically just a simple rename and publication of the already implemented functions. To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macroCONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.Also make sure the requested payload is aligned to full pages, otherwise this simple driver fails to load the last page.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>---
drivers/mtd/nand/mxc_nand_spl.c | 13 ++++++++++--- include/configs/mx31pdk.h | 1 + include/configs/tx25.h | 1 + 3 files changed, 12 insertions(+), 3 deletions(-)drivers/mtd/nand/mxc_nand_spl.c +1
-1@@ -290,7 +290,7 @@ static int is_badblock(int pagenumber)return 0 ; } -static int nand_load (unsigned int from,unsigned int size, unsigned char *buf) +int nand_spl_load_image(uint32_t from, unsigned int size, void*buf)
{
drivers/mtd/nand/mxc_nand_spl.c +1
-0int i;unsigned int page;@@ -303,6 +303,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) page = from /CONFIG_SYS_NAND_PAGE_SIZE ; i =0 ; + size =roundup (size,CONFIG_SYS_NAND_PAGE_SIZE );drivers/mtd/nand/mxc_nand_spl.c +1
-0while (i < size /CONFIG_SYS_NAND_PAGE_SIZE ) {if (nfc_read_page (page, buf) <0 )return -1 ;@@ -332,6 +333,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) return 0; } +#ifdef CONFIG_SPL_NAND_SUPPORT_LEGACY /*
* The main entry for NAND booting. It's necessary that SDRAM is already * configured and available since this code loads the main U-Boot image
drivers/mtd/nand/mxc_nand_spl.c +0
-1@@ -345,8 +347,9 @@ void nand_boot(void)* CONFIG_SYS_NAND_U_BOOT_OFFS and CONFIG_SYS_NAND_U_BOOT_SIZE must * be aligned to full pages */ - if (!nand_load(CONFIG_SYS_NAND_U_BOOT_OFFS,
CONFIG_SYS_NAND_U_BOOT_SIZE, - (uchar
*)CONFIG_SYS_NAND_U_BOOT_DST)) {drivers/mtd/nand/mxc_nand_spl.c +7
-0+if (!nand_spl_load_image (CONFIG_SYS_NAND_U_BOOT_OFFS , +CONFIG_SYS_NAND_U_BOOT_SIZE , + (uchar *)CONFIG_SYS_NAND_U_BOOT_DST )) {/* Copy from NAND successful, start U-boot */ uboot = (void *)CONFIG_SYS_NAND_U_BOOT_START ;uboot ();@@ -364,3 +367,7 @@ void hang(void)/* Loop forever */ while (1) ; } +#endif + +void nand_init (void ) {} +void nand_deselect (void ) {}(quoted diff) +0
-0
b/include/configs/mx31pdk.hindex 1754595..217552e 100644include/configs/mx31pdk.h +0
-0@@ -50,6 +50,7 @@Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
include/configs/mx31pdk.h line 51
Quoted 1 line
include/configs/mx31pdk.h +0
-0#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-
spl.lds"Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
include/configs/mx31pdk.h +1
-0#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x87dc0000
#defineCONFIG_SYS_TEXT_BASE0x87e00000Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
include/configs/tx25.h line 38
Quoted 7 lines
include/configs/tx25.h +0
-0@@ -37,6 +37,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-
Quoted 50 lines
Quoted 42 lines
Quoted 14 lines
spl.lds"Quoted 11 lines
Quoted 11 lines
Quoted 10 lines
include/configs/tx25.h +1
-0#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x810c0000
#defineCONFIG_SYS_TEXT_BASE0x81200000--
1.7.11.7Quoted 8 lines
Quoted 6 lines
Quoted 6 lines
This is not about legacy vs. non-legacy. This is about basic vs. more featured
SPL because of SPL size constraints. So what about droppingCONFIG_SPL_NAND_SUPPORT_LEGACYand testing forCONFIG_SPL_FRAMEWORKdefinition
instead?I was thinking about that, but the symbol is unrelated to NAND.
Well, it's
CONFIG_SPL_FRAMEWORK+CONFIG_SPL_NAND_SUPPORTthat defines the other
implementation, andCONFIG_SPL_NAND_SUPPORTtriggers the build ofmxc_nand_spl.c
for SPL, so the common point isCONFIG_SPL_FRAMEWORK.Quoted 4 lines
I still think
it's either a matter of fixing for new SPL or removing those two boards. The
nand_spl/ stuff shall be removed ASAP.Removing those boards is not a solution.
Is it really about "new" SPL? The generic SPL is enabled by
CONFIG_SPL, andCONFIG_SPL_FRAMEWORKis a sub-option. If the generic SPL rules imposed to use the SPL framework functions, there would be no such sub-option. So it looks like
these boards are complying to the new SPL rules.Quoted 7 lines
We could see if using
CONFIG_SPL_FRAMEWORKwould still allow the SPL to fit in
2 kiB in order to drop this function, but if it does not fit, the new SPL rules
should still make it possible to have a solution for any board having SPL size
constraints.Quoted 5 lines
I've just quickly build-tested the hack-patch below to have a guess of the SPL size for mx31pdk with
CONFIG_SPL_FRAMEWORK, and I get 3.9 kiB, so way more than the maximum 2 kiB. Hence, these boards can't unfortunately be made more generic, but it's not a reason to remove them. The generic SPL must allow such hardware to be supported by U-Boot.
Quoted 5 lines
See the other email for my other idea. Either we can go with that or I can focus
on making use of the regular MXC NAND driver on mx5, since there's no 2KiB limit
there.
The latter is probably overkill and not worth it.
Best regards,
Beno?t
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
In reply to Benoît ThébaudeauDear Beno?t Th?baudeau,
Quoted 173 lines
Dear Marek Vasut,
Quoted 146 lines · Marek Vasut
Dear Beno?t Th?baudeau,
Quoted 123 lines · Beno?t Th?baudeau
Quoted 122 lines
Dear Marek Vasut,
Quoted 119 lines · Marek Vasut
Add support for generic NAND SPL via the SPL framework into the
mxc_nand_spldriver. This is basically just a simple rename and publication of the already implemented functions. To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macroCONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.Also make sure the requested payload is aligned to full pages, otherwise this simple driver fails to load the last page.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>---
drivers/mtd/nand/mxc_nand_spl.c | 13 ++++++++++--- include/configs/mx31pdk.h | 1 + include/configs/tx25.h | 1 + 3 files changed, 12 insertions(+), 3 deletions(-)drivers/mtd/nand/mxc_nand_spl.c +1
-1@@ -290,7 +290,7 @@ static int is_badblock(int pagenumber)return 0 ; } -static int nand_load (unsigned int from,unsigned int size, unsigned char *buf) +int nand_spl_load_image(uint32_t from, unsigned int size, void*buf)
{
drivers/mtd/nand/mxc_nand_spl.c +1
-0int i;unsigned int page;@@ -303,6 +303,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) page = from /CONFIG_SYS_NAND_PAGE_SIZE ; i =0 ; + size =roundup (size,CONFIG_SYS_NAND_PAGE_SIZE );drivers/mtd/nand/mxc_nand_spl.c +1
-0while (i < size /CONFIG_SYS_NAND_PAGE_SIZE ) {if (nfc_read_page (page, buf) <0 )return -1 ;@@ -332,6 +333,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) return 0; } +#ifdef CONFIG_SPL_NAND_SUPPORT_LEGACY /*
* The main entry for NAND booting. It's necessary that SDRAM is already * configured and available since this code loads the main U-Boot image
drivers/mtd/nand/mxc_nand_spl.c +0
-1@@ -345,8 +347,9 @@ void nand_boot(void)* CONFIG_SYS_NAND_U_BOOT_OFFS and CONFIG_SYS_NAND_U_BOOT_SIZE must * be aligned to full pages */ - if (!nand_load(CONFIG_SYS_NAND_U_BOOT_OFFS,
CONFIG_SYS_NAND_U_BOOT_SIZE, - (uchar
*)CONFIG_SYS_NAND_U_BOOT_DST)) {drivers/mtd/nand/mxc_nand_spl.c +7
-0+if (!nand_spl_load_image (CONFIG_SYS_NAND_U_BOOT_OFFS , +CONFIG_SYS_NAND_U_BOOT_SIZE , + (uchar *)CONFIG_SYS_NAND_U_BOOT_DST )) {/* Copy from NAND successful, start U-boot */ uboot = (void *)CONFIG_SYS_NAND_U_BOOT_START ;uboot ();@@ -364,3 +367,7 @@ void hang(void)/* Loop forever */ while (1) ; } +#endif + +void nand_init (void ) {} +void nand_deselect (void ) {}include/configs/mx31pdk.h +1
-0@@ -50,6 +50,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-spl.lds"#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x87dc0000
#defineCONFIG_SYS_TEXT_BASE0x87e00000Quoted 7 lines
Quoted 7 lines
include/configs/tx25.h line 38
Quoted 7 lines
include/configs/tx25.h +0
-0@@ -37,6 +37,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-
spl.lds"Quoted 11 lines
Quoted 11 lines
Quoted 10 lines
include/configs/tx25.h +1
-0#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x810c0000
#defineCONFIG_SYS_TEXT_BASE0x81200000--
1.7.11.7Quoted 8 lines
Quoted 6 lines
Quoted 6 lines
This is not about legacy vs. non-legacy. This is about basic vs. more featured
SPL because of SPL size constraints. So what about droppingCONFIG_SPL_NAND_SUPPORT_LEGACYand testing forCONFIG_SPL_FRAMEWORKdefinition
instead?I was thinking about that, but the symbol is unrelated to NAND.
Well, it's
CONFIG_SPL_FRAMEWORK+CONFIG_SPL_NAND_SUPPORTthat defines the other implementation, andCONFIG_SPL_NAND_SUPPORTtriggers the build ofmxc_nand_spl.cfor SPL, so the common point isCONFIG_SPL_FRAMEWORK.Quoted 3 lines
I still think
it's either a matter of fixing for new SPL or removing those two boards. The nand_spl/ stuff shall be removed ASAP.Removing those boards is not a solution.
Is it really about "new" SPL? The generic SPL is enabled by
CONFIG_SPL, andCONFIG_SPL_FRAMEWORKis a sub-option. If the generic SPL rules imposed to use the SPL framework functions, there would be no such sub-option. So it looks like these boards are complying to the new SPL rules.
Quoted 4 lines
We could see if using
CONFIG_SPL_FRAMEWORKwould still allow the SPL to fit in 2 kiB in order to drop this function, but if it does not fit, the new SPL rules should still make it possible to have a solution for any board having SPL size constraints.
Rather than this, the other option would be to make whatever calls nand_boot() compatible with the SPL frameworks' implementation of spl_nand, so we don't need different function calls.
Quoted 9 lines
Quoted 4 lines
Quoted 2 lines
Also, I don't see any
nand_init()(called byspl_nand_load_image()) for m53evk.What do you mean?
I mean that I don't identify the implementation of
nand_init()used in the context of the m53evk since the builds ofdrivers/mtd/nand/nand.canddrivers/mtd/nand/nand_spl_simple.cdo not seem to be enabled for this board, and I don't see another definition.
It's there, as you said somewhere in the thread below ;-)
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
Dear Marek Vasut,
Quoted 182 lines · Marek Vasut
Subject: Re: [PATCH 2/6] nand: Add
SPL_NANDsupport tomxc_nand_splDear Beno?t Th?baudeau,
Quoted 173 lines
Dear Marek Vasut,
Quoted 146 lines · Marek Vasut
Dear Beno?t Th?baudeau,
Quoted 123 lines · Beno?t Th?baudeau
Quoted 122 lines
Dear Marek Vasut,
Quoted 119 lines · Marek Vasut
Add support for generic NAND SPL via the SPL framework into the
mxc_nand_spldriver. This is basically just a simple rename and publication of the already implemented functions. To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macroCONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.Also make sure the requested payload is aligned to full pages, otherwise this simple driver fails to load the last page.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>---
drivers/mtd/nand/mxc_nand_spl.c | 13 ++++++++++--- include/configs/mx31pdk.h | 1 + include/configs/tx25.h | 1 + 3 files changed, 12 insertions(+), 3 deletions(-)drivers/mtd/nand/mxc_nand_spl.c +1
-1@@ -290,7 +290,7 @@ static int is_badblock(int pagenumber)return 0 ; } -static int nand_load (unsigned int from,unsigned int size, unsigned char *buf) +int nand_spl_load_image(uint32_t from, unsigned int size, void*buf)
{
drivers/mtd/nand/mxc_nand_spl.c +1
-0int i;unsigned int page;@@ -303,6 +303,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) page = from /CONFIG_SYS_NAND_PAGE_SIZE ; i =0 ; + size =roundup (size,CONFIG_SYS_NAND_PAGE_SIZE );drivers/mtd/nand/mxc_nand_spl.c +1
-0while (i < size /CONFIG_SYS_NAND_PAGE_SIZE ) {if (nfc_read_page (page, buf) <0 )return -1 ;@@ -332,6 +333,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) return 0; } +#ifdef CONFIG_SPL_NAND_SUPPORT_LEGACY /*
* The main entry for NAND booting. It's necessary that SDRAM is already * configured and available since this code loads the main U-Boot image
drivers/mtd/nand/mxc_nand_spl.c +0
-1@@ -345,8 +347,9 @@ void nand_boot(void)* CONFIG_SYS_NAND_U_BOOT_OFFS and CONFIG_SYS_NAND_U_BOOT_SIZE must * be aligned to full pages */ - if (!nand_load(CONFIG_SYS_NAND_U_BOOT_OFFS,
CONFIG_SYS_NAND_U_BOOT_SIZE, - (uchar
*)CONFIG_SYS_NAND_U_BOOT_DST)) {drivers/mtd/nand/mxc_nand_spl.c +7
-0+if (!nand_spl_load_image (CONFIG_SYS_NAND_U_BOOT_OFFS , +CONFIG_SYS_NAND_U_BOOT_SIZE , + (uchar *)CONFIG_SYS_NAND_U_BOOT_DST )) {/* Copy from NAND successful, start U-boot */ uboot = (void *)CONFIG_SYS_NAND_U_BOOT_START ;uboot ();@@ -364,3 +367,7 @@ void hang(void)/* Loop forever */ while (1) ; } +#endif + +void nand_init (void ) {} +void nand_deselect (void ) {}include/configs/mx31pdk.h +1
-0@@ -50,6 +50,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-spl.lds"#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x87dc0000
#defineCONFIG_SYS_TEXT_BASE0x87e00000Quoted 7 lines
Quoted 7 lines
include/configs/tx25.h line 38
Quoted 7 lines
include/configs/tx25.h +0
-0@@ -37,6 +37,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-
spl.lds"Quoted 11 lines
Quoted 11 lines
Quoted 10 lines
include/configs/tx25.h +1
-0#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x810c0000
#defineCONFIG_SYS_TEXT_BASE0x81200000--
1.7.11.7Quoted 8 lines
Quoted 6 lines
Quoted 6 lines
This is not about legacy vs. non-legacy. This is about basic vs. more featured
SPL because of SPL size constraints. So what about droppingCONFIG_SPL_NAND_SUPPORT_LEGACYand testing forCONFIG_SPL_FRAMEWORKdefinition
instead?I was thinking about that, but the symbol is unrelated to NAND.
Well, it's
CONFIG_SPL_FRAMEWORK+CONFIG_SPL_NAND_SUPPORTthat defines the other implementation, andCONFIG_SPL_NAND_SUPPORTtriggers the build ofmxc_nand_spl.cfor SPL, so the common point isCONFIG_SPL_FRAMEWORK.Quoted 3 lines
I still think
it's either a matter of fixing for new SPL or removing those two boards. The nand_spl/ stuff shall be removed ASAP.Removing those boards is not a solution.
Is it really about "new" SPL? The generic SPL is enabled by
CONFIG_SPL, andCONFIG_SPL_FRAMEWORKis a sub-option. If the generic SPL rules imposed to use the SPL framework functions, there would be no such sub-option. So it looks like these boards are complying to the new SPL rules.Quoted 4 lines
We could see if using
CONFIG_SPL_FRAMEWORKwould still allow the SPL to fit in 2 kiB in order to drop this function, but if it does not fit, the new SPL rules should still make it possible to have a solution for any board having SPL size constraints.
Quoted 5 lines
Rather than this, the other option would be to make whatever calls
nand_boot()
compatible with the SPL frameworks' implementation ofspl_nand, so we don't need
different function calls.
If by that you mean renaming and reorganizing functions without using the main SPL framework, sure. For hang(), this has already been done by Andreas' pending series, and for nand_boot(), Tom's plan seems to be to merge the various existing implementations, and beyond that it would be possible to make the calls look more like the SPL framework's as you suggest.
Best regards,
Beno?t
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
Dear Beno?t Th?baudeau,
Quoted 126 lines
Dear Marek Vasut,
Quoted 123 lines · Marek Vasut
Subject: Re: [PATCH 2/6] nand: Add
SPL_NANDsupport tomxc_nand_splDear Beno?t Th?baudeau,
Quoted 119 lines
Dear Marek Vasut,
Quoted 116 lines · Marek Vasut
Dear Beno?t Th?baudeau,
Quoted 114 lines · Beno?t Th?baudeau
Quoted 113 lines
Dear Marek Vasut,
Quoted 110 lines · Marek Vasut
Add support for generic NAND SPL via the SPL framework into the
mxc_nand_spldriver. This is basically just a simple rename and publication of the already implemented functions. To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macroCONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.Also make sure the requested payload is aligned to full pages, otherwise this simple driver fails to load the last page.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>---
drivers/mtd/nand/mxc_nand_spl.c | 13 ++++++++++--- include/configs/mx31pdk.h | 1 + include/configs/tx25.h | 1 + 3 files changed, 12 insertions(+), 3 deletions(-)drivers/mtd/nand/mxc_nand_spl.c +1
-1@@ -290,7 +290,7 @@ static int is_badblock(int pagenumber)return 0 ; } -static int nand_load (unsigned int from,unsigned int size, unsigned char *buf) +int nand_spl_load_image(uint32_t from, unsigned int size, void*buf)
{
drivers/mtd/nand/mxc_nand_spl.c +1
-0int i;unsigned int page;@@ -303,6 +303,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) page = from /CONFIG_SYS_NAND_PAGE_SIZE ; i =0 ; + size =roundup (size,CONFIG_SYS_NAND_PAGE_SIZE );drivers/mtd/nand/mxc_nand_spl.c +1
-0while (i < size /CONFIG_SYS_NAND_PAGE_SIZE ) {if (nfc_read_page (page, buf) <0 )return -1 ;@@ -332,6 +333,7 @@ static int nand_load(unsigned int from, unsigned int size, unsigned char *buf) return 0; } +#ifdef CONFIG_SPL_NAND_SUPPORT_LEGACY /*
* The main entry for NAND booting. It's necessary that SDRAM is already * configured and available since this code loads the main U-Boot image
drivers/mtd/nand/mxc_nand_spl.c +0
-1@@ -345,8 +347,9 @@ void nand_boot(void)* CONFIG_SYS_NAND_U_BOOT_OFFS and CONFIG_SYS_NAND_U_BOOT_SIZE must * be aligned to full pages */ - if (!nand_load(CONFIG_SYS_NAND_U_BOOT_OFFS,
CONFIG_SYS_NAND_U_BOOT_SIZE, - (uchar
*)CONFIG_SYS_NAND_U_BOOT_DST)) {drivers/mtd/nand/mxc_nand_spl.c +7
-0+if (!nand_spl_load_image (CONFIG_SYS_NAND_U_BOOT_OFFS , +CONFIG_SYS_NAND_U_BOOT_SIZE , + (uchar *)CONFIG_SYS_NAND_U_BOOT_DST )) {/* Copy from NAND successful, start U-boot */ uboot = (void *)CONFIG_SYS_NAND_U_BOOT_START ;uboot ();@@ -364,3 +367,7 @@ void hang(void)/* Loop forever */ while (1) ; } +#endif + +void nand_init (void ) {} +void nand_deselect (void ) {}(quoted diff) +0
-0
b/include/configs/mx31pdk.hindex 1754595..217552e 100644include/configs/mx31pdk.h +0
-0@@ -50,6 +50,7 @@
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
include/configs/mx31pdk.h line 51
Quoted 1 line
include/configs/mx31pdk.h +0
-0#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-boot-
spl.lds"
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
include/configs/mx31pdk.h +1
-0#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x87dc0000
#defineCONFIG_SYS_TEXT_BASE0x87e00000
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
include/configs/tx25.h line 38
Quoted 7 lines
include/configs/tx25.h +0
-0@@ -37,6 +37,7 @@#define CONFIG_SPL_LDSCRIPT "arch/$(ARCH)/cpu/u-
boot-
Quoted 51 lines
Quoted 45 lines
Quoted 40 lines
Quoted 14 lines
spl.lds"Quoted 11 lines
Quoted 11 lines
Quoted 10 lines
include/configs/tx25.h +1
-0#define CONFIG_SPL_MAX_SIZE 2048#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SPL_NAND_SUPPORT_LEGACY #define
CONFIG_SPL_TEXT_BASE0x810c0000
#defineCONFIG_SYS_TEXT_BASE0x81200000--
1.7.11.7Quoted 8 lines
Quoted 6 lines
Quoted 6 lines
This is not about legacy vs. non-legacy. This is about basic vs. more featured
SPL because of SPL size constraints. So what about droppingCONFIG_SPL_NAND_SUPPORT_LEGACYand testing forCONFIG_SPL_FRAMEWORKdefinition
instead?I was thinking about that, but the symbol is unrelated to NAND.
Well, it's
CONFIG_SPL_FRAMEWORK+CONFIG_SPL_NAND_SUPPORTthat defines the other implementation, andCONFIG_SPL_NAND_SUPPORTtriggers the build ofmxc_nand_spl.cfor SPL, so the common point isCONFIG_SPL_FRAMEWORK.Quoted 3 lines
I still think
it's either a matter of fixing for new SPL or removing those two boards. The nand_spl/ stuff shall be removed ASAP.Removing those boards is not a solution.
Is it really about "new" SPL? The generic SPL is enabled by
CONFIG_SPL, andCONFIG_SPL_FRAMEWORKis a sub-option. If the generic SPL rules imposed to use the SPL framework functions, there would be no such sub-option. So it looks like these boards are complying to the new SPL rules.Quoted 4 lines
We could see if using
CONFIG_SPL_FRAMEWORKwould still allow the SPL to fit in 2 kiB in order to drop this function, but if it does not fit, the new SPL rules should still make it possible to have a solution for any board having SPL size constraints.Quoted 5 lines
Rather than this, the other option would be to make whatever calls
nand_boot()
compatible with the SPL frameworks' implementation ofspl_nand, so we don't need
different function calls.
Quoted 6 lines
If by that you mean renaming and reorganizing functions without using the main SPL framework, sure. For
hang(), this has already been done by Andreas' pending series, and fornand_boot(), Tom's plan seems to be to merge the various existing implementations, and beyond that it would be possible to make the calls look more like the SPL framework's as you suggest.
OK, makes sense.
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
In reply to Marek VasutQuoted 5 lines · Marek Vasut
To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macro
CONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.
It sounds like "fixing" these boards is mainly a matter of confirming that a configuration for them based around CONFIG_SPL_FRAMEWORK will fit in 2K. If so, I don't think there is any reason to keep legacy support around.
drivers/mtd/nand/mxc_nand_spl.c void hang(void) · line 373
Quoted 2 lines
drivers/mtd/nand/mxc_nand_spl.c +2
-0+void nand_init (void ) {} +void nand_deselect (void ) {}
Couldn't you just rename nfc_nand_init() to nand_init()?
- Philip
Signature · 3 lines
Philip Paeps Senior Reality Engineer Ministry of Information
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
Quoted 11 lines · Philip Paeps
Quoted 5 lines · Marek Vasut
To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macro
CONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.It sounds like "fixing" these boards is mainly a matter of confirming that a configuration for them based around
CONFIG_SPL_FRAMEWORKwill fit in 2K. If so, I don't think there is any reason to keep legacy support around.
A first build with CONFIG_SPL_FRAMEWORK came out to nearly 4K. Large contributors being (unsurprisingly) libcommon and libgeneric. I had to get rid of a puts() in libspl to make it build without those libraries. Unfortunately, that still came out to 2.2K. Close. :-)
I couldn't identify any obvious 100 bytes to scrap from glancing at u-boot-spl.map or objdump -D u-boot-spl, but I'll take a look.
- Philip
common/spl/spl.c +0 -1
@@ -50,7 +50,6 @@ static bd_t bdata __attribute__ ((section(".data")));inline void hang (void ) { -puts ("### ERROR ### Please RESET the board ###\n" );for (;;) ; }
Signature · 3 lines
Philip Paeps Senior Reality Engineer Ministry of Information
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
Hi Philip,
Quoted 13 lines · Philip Paeps
Quoted 11 lines · Philip Paeps
Quoted 5 lines · Marek Vasut
To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macro
CONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.It sounds like "fixing" these boards is mainly a matter of confirming that a configuration for them based around
CONFIG_SPL_FRAMEWORKwill fit in 2K. If so, I don't think there is any reason to keep legacy support around.
Quoted 7 lines
A first build with
CONFIG_SPL_FRAMEWORKcame out to nearly 4K. Large contributors being (unsurprisingly) libcommon and libgeneric. I had to get rid of aputs()in libspl to make it build without those libraries. Unfortunately, that still came out to 2.2K. Close. :-)I couldn't identify any obvious 100 bytes to scrap from glancing at u-boot-spl.map or objdump -D u-boot-spl, but I'll take a look.
Cool. But don't forget that the build must also succeed whatever the toolchain, so there must be a margin on the 2 kiB, perhaps 5% to 10%.
Another way around could be to just move nand_boot() to a more generic SPL location so that other hardware having minimal SPL size requirements can use it without duplicating it. As to hang(), it has already been handled by Andreas.
Best regards,
Beno?t
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
In reply to Philip PaepsQuoted 13 lines · Philip Paeps
Quoted 11 lines · Philip Paeps
Quoted 5 lines · Marek Vasut
To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macro
CONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.It sounds like "fixing" these boards is mainly a matter of confirming that a configuration for them based around
CONFIG_SPL_FRAMEWORKwill fit in 2K. If so, I don't think there is any reason to keep legacy support around.
Quoted 7 lines
A first build with
CONFIG_SPL_FRAMEWORKcame out to nearly 4K. Large contributors being (unsurprisingly) libcommon and libgeneric. I had to get rid of aputs()in libspl to make it build without those libraries. Unfortunately, that still came out to 2.2K. Close. :-)I couldn't identify any obvious 100 bytes to scrap from glancing at u-boot-spl.map or objdump -D u-boot-spl, but I'll take a look.
Just as I hit 'send', it occurred to me that this configuration is with a fairly lengthy lowlevel_init.S to support external boot. Paring that to the bare minimum gives a u-boot-spl.bin of 1821 bytes.
I'm not familiar enough with 'internal boot' and the 'imximage' format to judge whether this leaves enough margin for telling the ROM loader to do its thing usefully. Beno?t: do you have an idea how long a typically useful imximage preamble gets? Is ~230 bytes sufficient margin?
- Philip
Signature · 3 lines
Philip Paeps Senior Reality Engineer Ministry of Information
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
Hi Philip,
Quoted 19 lines · On Friday, April 19, 2013 5:09:59 PM,Philip Paeps wrote:
Quoted 18 lines · Philip Paeps
Quoted 11 lines · Philip Paeps
Quoted 5 lines · Marek Vasut
To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macro
CONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.It sounds like "fixing" these boards is mainly a matter of confirming that a configuration for them based around
CONFIG_SPL_FRAMEWORKwill fit in 2K. If so, I don't think there is any reason to keep legacy support around.A first build with
CONFIG_SPL_FRAMEWORKcame out to nearly 4K. Large contributors being (unsurprisingly) libcommon and libgeneric. I had to get rid of aputs()in libspl to make it build without those libraries. Unfortunately, that still came out to 2.2K. Close. :-)
Quoted 6 lines
Quoted 2 lines
I couldn't identify any obvious 100 bytes to scrap from glancing at u-boot-spl.map or objdump -D u-boot-spl, but I'll take a look.
Just as I hit 'send', it occurred to me that this configuration is with a fairly lengthy
lowlevel_init.Sto support external boot. Paring that to the bare minimum gives a u-boot-spl.bin of 1821 bytes.
But this requires a board-specific lowlevel_init() and a hack for puts() (which is perhaps already solved by Andreas' series), just to bloat the SPL with stuff useless for those boards, vs. a simple nand_boot() that can be made common to all SPLs with size restrictions.
Quoted 4 lines
I'm not familiar enough with 'internal boot' and the 'imximage' format to judge whether this leaves enough margin for telling the ROM loader to do its thing usefully. Beno?t: do you have an idea how long a typically useful imximage preamble gets? Is ~230 bytes sufficient margin?
What do you mean? mx31pdk and tx25, contrary to m53evk, use external NAND boot, hence the 2-kiB SPL size limit (which could actually be extended to 4 kiB for tx25, but not for mx31pdk), and they don't require any imximage.
Best regards,
Beno?t
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
Quoted 30 lines · Beno?t Th?baudeau
Hi Philip,
Quoted 20 lines · On Friday, April 19, 2013 5:09:59 PM,Philip Paeps wrote:
Quoted 19 lines · Philip Paeps
Quoted 7 lines · Philip Paeps
Quoted 5 lines · Marek Vasut
To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macro
CONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.Quoted 5 lines
It sounds like "fixing" these boards is mainly a matter of confirming that a configuration for them based around
CONFIG_SPL_FRAMEWORKwill fit
in 2K. If so, I don't think there is any reason to keep legacy support around.A first build with
CONFIG_SPL_FRAMEWORKcame out to nearly 4K. Large contributors being (unsurprisingly) libcommon and libgeneric. I had to get rid of aputs()in libspl to make it build without those libraries. Unfortunately, that still came out to 2.2K. Close. :-)Quoted 6 lines
Quoted 2 lines
I couldn't identify any obvious 100 bytes to scrap from glancing at u-boot-spl.map or objdump -D u-boot-spl, but I'll take a look.
Just as I hit 'send', it occurred to me that this configuration is with a fairly lengthy
lowlevel_init.Sto support external boot. Paring that to the bare minimum gives a u-boot-spl.bin of 1821 bytes.
Quoted 6 lines
But this requires a board-specific
lowlevel_init()and a hack forputs()(which
is perhaps already solved by Andreas' series), just to bloat the SPL with stuff useless for those boards, vs. a simplenand_boot()that can be made common
to all SPLs with size restrictions.
Or, with Andreas' series to get rid of puts(), and with an option to make the use of spl_parse_image_header() an option (i.e. to be able to force raw binary SPL payload format), the change would be almost equivalent to using the current nand_boot(), so the size should follow.
Best regards,
Beno?t
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
In reply to Benoît ThébaudeauQuoted 10 lines · Beno?t Th?baudeau
Quoted 9 lines · On Friday, April 19, 2013 5:09:59 PM,Philip Paeps wrote:
Quoted 7 lines · Philip Paeps
A first build with
CONFIG_SPL_FRAMEWORKcame out to nearly 4K. Large contributors being (unsurprisingly) libcommon and libgeneric. I had to get rid of aputs()in libspl to make it build without those libraries. Unfortunately, that still came out to 2.2K. Close. :-)I couldn't identify any obvious 100 bytes to scrap from glancing at u-boot-spl.map or objdump -D u-boot-spl, but I'll take a look.
Quoted 8 lines
Quoted 3 lines
Just as I hit 'send', it occurred to me that this configuration is with a fairly lengthy
lowlevel_init.Sto support external boot. Paring that to the bare minimum gives a u-boot-spl.bin of 1821 bytes.But this requires a board-specific
lowlevel_init()and a hack forputs()(which is perhaps already solved by Andreas' series), just to bloat the SPL with stuff useless for those boards, vs. a simplenand_boot()that can be made common to all SPLs with size restrictions.
Comparing the contents of "framework" SPL with "old" SPL, it looks like the only thing we gain is support for booting uImage files. At the cost of significantly reduced room for flexibility in lowlevel_init().
I agree that a simple nand_boot() is probably the way forward.
Quoted 8 lines
Quoted 4 lines
I'm not familiar enough with 'internal boot' and the 'imximage' format to judge whether this leaves enough margin for telling the ROM loader to do its thing usefully. Beno?t: do you have an idea how long a typically useful imximage preamble gets? Is ~230 bytes sufficient margin?
What do you mean? mx31pdk and tx25, contrary to m53evk, use external NAND boot, hence the 2-kiB SPL size limit (which could actually be extended to 4 kiB for tx25, but not for mx31pdk), and they don't require any imximage.
Mmm. Oh. It looks like I was looking at the wrong lowlevel_init() functions. They could be made to work with a very stripped down lowlevel_init(), but it's a _very_ tight fit.
- Philip
Signature · 3 lines
Philip Paeps Senior Reality Engineer Ministry of Information
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Quoted 26 lines · Philip Paeps
Quoted 18 lines · Beno?t Th?baudeau
Quoted 8 lines · On Friday, April 19, 2013 5:09:59 PM,Philip Paeps wrote:
Quoted 6 lines · Philip Paeps
A first build with
CONFIG_SPL_FRAMEWORKcame out to nearly 4K. Large contributors being (unsurprisingly) libcommon and libgeneric. I had to get rid of aputs()in libspl to make it build without those libraries. Unfortunately, that still came out to 2.2K. Close. :-)Quoted 8 lines
Quoted 3 lines
I couldn't identify any obvious 100 bytes to scrap from glancing at u-boot-spl.map or objdump -D u-boot-spl, but I'll take a look.
Just as I hit 'send', it occurred to me that this configuration is with a fairly lengthy
lowlevel_init.Sto support external boot. Paring that to the bare minimum gives a u-boot-spl.bin of 1821 bytes.Quoted 5 lines
But this requires a board-specific
lowlevel_init()and a hack forputs()(which is perhaps already solved by Andreas' series), just to bloat the SPL with stuff useless for those boards, vs. a simplenand_boot()that can be made common to all SPLs with size restrictions.
Quoted 6 lines
Comparing the contents of "framework" SPL with "old" SPL, it looks like the only thing we gain is support for booting uImage files. At the cost of significantly reduced room for flexibility in
lowlevel_init().I agree that a simple
nand_boot()is probably the way forward.
Note that PowerPC has had to deal with this as well, see drivers/mtd/nand/fsl_elbc_spl.c and I suspect some amount of that should be made more easily borrowable for the case where SPL doesn't (or can't/won't) deal with image headers.
- --
Tom
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
iQIcBAEBAgAGBQJRcW7JAAoJENk4IS6UOR1Wmx4P/iIJS8QXZfdO7FNpeZz7Zet6 UNwsFoZeZ1jq+7DhFoA+o5Vwafys+AcYXX2uP0pVC9TJAbGBGSprYao6NE8tA0oo OWFWktUHoNn5vPedcrZuN6fxcBT9p+7MCo76YojSnWWd2DuwODODY6G3MX7I9iMi qbsBxJ9R13ubvd3rXGLltDFmR3hZx6wNRv/ke7Mk1dTycEKFbQIFmYhyXLHJLB6+ wxon4knbYcz1GmISUioGDX+hmc4NmKapydSeolmN1wE2tY9+0Mw+OINXcrYTjAN+ 8HpeuDsY2d/MzFdfgf9XvrZh2BFWmT7infK09udd6mbxosQYmEc8jIDnk+3DtA1I blcLjHOkxdVBS1665D0rWVGHs41CQcJfaaSdb3+pZSAbRwNzh2vivuLwdI//LyHf oMivbM8fLWkSpMSAxu517rfutF+KiJRr0Bns/QqQn4JmIs3qLfjPvpdPST/BaaI9 CAzlpz9c+iZaFvmfutvXhkPOZOKsjAfvT/04SAdXiruf83tn5ceU3/Z/adVxycth i8UdVqjAG4wtZtGaGjwd7SsW5cFlNy0ODZHMY+QqxD+XSD6T+3L815QpPZDhM2T2 obhgujYC6gg6vAlFp5nWnRxjeL0vB5k/HqV+4k9V9dAcpOyhWRvLeKgaiGyQMSpL xTyf7qQrf4ZkGFjO1zU/
=lMuD
-----END PGP SIGNATURE-----
PATCH 2/6 nand: Add SPL_NAND support to mxc_nand_spl
In reply to Philip PaepsDear Philip Paeps,
Quoted 11 lines · Marek Vasut
Quoted 5 lines
To avoid the old function which are used with the nand_spl/ stuff getting in the way of NAND SPL framework, the macro
CONFIG_SPL_NAND_LEGACYwas introduced and two remaining legacy boards were adjusted. These board need to be either fixed or removed in the long run, but I don't have either.It sounds like "fixing" these boards is mainly a matter of confirming that a configuration for them based around
CONFIG_SPL_FRAMEWORKwill fit in 2K. If so, I don't think there is any reason to keep legacy support around.
These boards are too limited, so we can do nothing about them but to keep this hack in.
Quoted 4 lines
drivers/mtd/nand/mxc_nand_spl.c void hang(void) · line 373
Quoted 2 lines
drivers/mtd/nand/mxc_nand_spl.c +2
-0+void nand_init (void ) {} +void nand_deselect (void ) {}Couldn't you just rename
nfc_nand_init()tonand_init()?
What good would that do?
Best regards,
Marek Vasut
PATCH 3/6 arm: imx: Fix u-boot-with-nand-spl.imx target
In reply to Marek VasutThis target is currently concatenating u-boot SPL in imximage format with u-boot.bin. The NAND SPL can load a raw binary, but the preffered format with much less limitations is uImage format. Fix the target so u-boot.bin is first converted into uImage format and only after that is concatenated.
--- arch/arm/imx-common/Makefile | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)
arch/arm/imx-common/Makefile +5 -2
@@ -69,8 +69,11 @@ $(OBJTREE)/u-boot-with-nand-spl.imx: $(OBJTREE)/SPL $(OBJTREE)/u-boot.bin -I binary -O binary $(OBJTREE)/spl/u-boot-nand-spl.imx \ $(OBJTREE)/spl/u-boot-nand-spl-pad.imx rm $(OBJTREE)/spl/u-boot-nand-spl.imx - cat $(OBJTREE)/spl/u-boot-nand-spl-pad.imx $(OBJTREE)/u-boot.bin > $@ - rm $(OBJTREE)/spl/u-boot-nand-spl-pad.imx + $(OBJTREE)/tools/mkimage -A arm -O U-Boot -a $(CONFIG_SYS_TEXT_BASE) \ + -e $(CONFIG_SYS_TEXT_BASE) -C none -d $(OBJTREE)/u-boot.bin \ + $(OBJTREE)/u-boot.uim + cat $(OBJTREE)/spl/u-boot-nand-spl-pad.imx $(OBJTREE)/u-boot.uim > $@ + rm $(OBJTREE)/spl/u-boot-nand-spl-pad.imx $(OBJTREE)/u-boot.uim #########################################################################
Signature · 1 line
1.7.11.7
PATCH 3/6 arm: imx: Fix u-boot-with-nand-spl.imx target
Dear Marek Vasut,
Quoted 5 lines · Marek Vasut
This target is currently concatenating u-boot SPL in imximage format with u-boot.bin. The NAND SPL can load a raw binary, but the preffered format with much less limitations is uImage format. Fix the target so u-boot.bin is first converted into uImage format and only after that is concatenated.
Please reword: This is not a fix, but an enhancement.
Quoted 32 lines
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>--- arch/arm/imx-common/Makefile | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)arch/arm/imx-common/Makefile +0
-0@@ -69,8 +69,11 @@ $(OBJTREE)/u-boot-with-nand-spl.imx: $(OBJTREE)/SPL$(OBJTREE)/u-boot.bin
arch/arm/imx-common/Makefile +5
-3-I binary -O binary $(OBJTREE)/spl/u-boot-nand-spl.imx \ $(OBJTREE)/spl/u-boot-nand-spl-pad.imx rm $(OBJTREE)/spl/u-boot-nand-spl.imx - cat $(OBJTREE)/spl/u-boot-nand-spl-pad.imx $(OBJTREE)/u-boot.bin > $@ - rm $(OBJTREE)/spl/u-boot-nand-spl-pad.imx + $(OBJTREE)/tools/mkimage -A arm -O U-Boot -a $(CONFIG_SYS_TEXT_BASE) \ + -e $(CONFIG_SYS_TEXT_BASE) -C none -d $(OBJTREE)/u-boot.bin \ + $(OBJTREE)/u-boot.uim + cat $(OBJTREE)/spl/u-boot-nand-spl-pad.imx $(OBJTREE)/u-boot.uim > $@ + rm $(OBJTREE)/spl/u-boot-nand-spl-pad.imx $(OBJTREE)/u-boot.uim ######################################################################### --
Quoted 1 line
1.7.11.7
This change is correct.
I don't know the uImage format very well. Can you give examples of what this patch can bring compared to the raw binary format?
If this change is useful, it should probably be made for u-boot-with-spl.imx too.
Best regards,
Beno?t
PATCH 3/6 arm: imx: Fix u-boot-with-nand-spl.imx target
Dear Beno?t Th?baudeau,
Quoted 51 lines
Dear Marek Vasut,
Quoted 5 lines · Marek Vasut
This target is currently concatenating u-boot SPL in imximage format with u-boot.bin. The NAND SPL can load a raw binary, but the preffered format with much less limitations is uImage format. Fix the target so u-boot.bin is first converted into uImage format and only after that is concatenated.
Please reword: This is not a fix, but an enhancement.
Quoted 36 lines
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>---
arch/arm/imx-common/Makefile | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)arch/arm/imx-common/Makefile +0
-0@@ -69,8 +69,11 @@ $(OBJTREE)/u-boot-with-nand-spl.imx: $(OBJTREE)/SPL$(OBJTREE)/u-boot.bin
arch/arm/imx-common/Makefile +5
-3-I binary -O binary $(OBJTREE)/spl/u-boot-nand-spl.imx \ $(OBJTREE)/spl/u-boot-nand-spl-pad.imx rm $(OBJTREE)/spl/u-boot-nand-spl.imx - cat $(OBJTREE)/spl/u-boot-nand-spl-pad.imx $(OBJTREE)/u-boot.bin > $@ - rm $(OBJTREE)/spl/u-boot-nand-spl-pad.imx + $(OBJTREE)/tools/mkimage -A arm -O U-Boot -a $(CONFIG_SYS_TEXT_BASE) \ + -e $(CONFIG_SYS_TEXT_BASE) -C none -d $(OBJTREE)/u-boot.bin \ + $(OBJTREE)/u-boot.uim + cat $(OBJTREE)/spl/u-boot-nand-spl-pad.imx $(OBJTREE)/u-boot.uim > $@ + rm $(OBJTREE)/spl/u-boot-nand-spl-pad.imx $(OBJTREE)/u-boot.uim ######################################################################## # --Quoted 1 line
1.7.11.7
This change is correct.
Quoted 5 lines
I don't know the uImage format very well. Can you give examples of what this patch can bring compared to the raw binary format?
If this change is useful, it should probably be made for u-boot-with-spl.imx too.
Please see common/spl/spl_nand.c which calls spl_parse_image_header() on the payload loaded from NAND. This is implemented in common/spl/spl.c and if proper header is not found, the payload is assumed to be of 200kB of size, which is far less than fully configured U-Boot. So using the uimage format is much more flexible.
Best regards,
Marek Vasut
PATCH 3/6 arm: imx: Fix u-boot-with-nand-spl.imx target
Dear Marek Vasut,
Quoted 59 lines · Marek Vasut
Dear Beno?t Th?baudeau,
Quoted 51 lines
Dear Marek Vasut,
Quoted 5 lines · Marek Vasut
This target is currently concatenating u-boot SPL in imximage format with u-boot.bin. The NAND SPL can load a raw binary, but the preffered format with much less limitations is uImage format. Fix the target so u-boot.bin is first converted into uImage format and only after that is concatenated.
Please reword: This is not a fix, but an enhancement.
Quoted 36 lines
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>---
arch/arm/imx-common/Makefile | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)arch/arm/imx-common/Makefile +0
-0@@ -69,8 +69,11 @@ $(OBJTREE)/u-boot-with-nand-spl.imx: $(OBJTREE)/SPL$(OBJTREE)/u-boot.bin
arch/arm/imx-common/Makefile +5
-3-I binary -O binary $(OBJTREE)/spl/u-boot-nand-spl.imx \ $(OBJTREE)/spl/u-boot-nand-spl-pad.imx rm $(OBJTREE)/spl/u-boot-nand-spl.imx - cat $(OBJTREE)/spl/u-boot-nand-spl-pad.imx $(OBJTREE)/u-boot.bin > $@ - rm $(OBJTREE)/spl/u-boot-nand-spl-pad.imx + $(OBJTREE)/tools/mkimage -A arm -O U-Boot -a $(CONFIG_SYS_TEXT_BASE) \ + -e $(CONFIG_SYS_TEXT_BASE) -C none -d $(OBJTREE)/u-boot.bin \ + $(OBJTREE)/u-boot.uim + cat $(OBJTREE)/spl/u-boot-nand-spl-pad.imx $(OBJTREE)/u-boot.uim > $@ + rm $(OBJTREE)/spl/u-boot-nand-spl-pad.imx $(OBJTREE)/u-boot.uim ######################################################################## # --Quoted 1 line
1.7.11.7
This change is correct.
Quoted 5 lines
I don't know the uImage format very well. Can you give examples of what this patch can bring compared to the raw binary format?
If this change is useful, it should probably be made for u-boot-with-spl.imx too.
Quoted 6 lines
Please see
common/spl/spl_nand.cwhich callsspl_parse_image_header()on the payload loaded from NAND. This is implemented incommon/spl/spl.cand if proper
header is not found, the payload is assumed to be of 200kB of size, which is far
less than fully configured U-Boot.
Or you could #define CONFIG_SYS_MONITOR_LEN to what you need.
Quoted 2 lines
So using the uimage format is much more
flexible.
I agree.
So I'd see the following changes for this patch:
- Reword "fix" to show enhancement, and detail the flexibility rationale in the description.
- Extend to u-boot-with-spl.imx.
Best regards,
Beno?t
PATCH 3/6 arm: imx: Fix u-boot-with-nand-spl.imx target
Dear Beno?t Th?baudeau,
Quoted 38 lines
Dear Marek Vasut,
Quoted 35 lines · Marek Vasut
Dear Beno?t Th?baudeau,
Quoted 33 lines
Dear Marek Vasut,
Quoted 5 lines · Marek Vasut
This target is currently concatenating u-boot SPL in imximage format with u-boot.bin. The NAND SPL can load a raw binary, but the preffered format with much less limitations is uImage format. Fix the target so u-boot.bin is first converted into uImage format and only after that is concatenated.
Please reword: This is not a fix, but an enhancement.
Quoted 22 lines
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>---
arch/arm/imx-common/Makefile | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)(quoted diff) +0
-0b/arch/arm/imx-common/Makefile index 44b6822..ba31d3e 100644
arch/arm/imx-common/Makefile +0
-0@@ -69,8 +69,11 @@ $(OBJTREE)/u-boot-with-nand-spl.imx:$(OBJTREE)/SPL $(OBJTREE)/u-boot.bin
arch/arm/imx-common/Makefile +0
-0-I binary -O binary $(OBJTREE)/spl/u-boot-nand-spl.imx \ $(OBJTREE)/spl/u-boot-nand-spl-pad.imx
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
arch/arm/imx-common/Makefile +0
-1rm $(OBJTREE)/spl/u-boot-nand-spl.imx - cat $(OBJTREE)/spl/u-boot-nand-spl-pad.imx $(OBJTREE)/u-boot.bin$@ - rm $(OBJTREE)/spl/u-boot-nand-spl-pad.imx
arch/arm/imx-common/Makefile +1
-0+ $(OBJTREE)/tools/mkimage -A arm -O U-Boot -a$(CONFIG_SYS_TEXT_BASE) \ + -e $(CONFIG_SYS_TEXT_BASE) -C
none -d
Quoted 19 lines
Quoted 19 lines
Quoted 18 lines
Quoted 10 lines
$(OBJTREE)/u-boot.bin \ + $(OBJTREE)/u-boot.uim
arch/arm/imx-common/Makefile +1
-0+ cat $(OBJTREE)/spl/u-boot-nand-spl-pad.imx $(OBJTREE)/u-boot.uim$@ + rm $(OBJTREE)/spl/u-boot-nand-spl-pad.imx $(OBJTREE)/u-boot.uim
####################################################################
#### #--
1.7.11.7This change is correct.
I don't know the uImage format very well. Can you give examples of what this patch can bring compared to the raw binary format?
If this change is useful, it should probably be made for u-boot-with-spl.imx too.
Quoted 8 lines
Quoted 6 lines
Please see
common/spl/spl_nand.cwhich callsspl_parse_image_header()on the payload loaded from NAND. This is implemented incommon/spl/spl.cand if proper
header is not found, the payload is assumed to be of 200kB of size, which is far
less than fully configured U-Boot.Or you could #define
CONFIG_SYS_MONITOR_LENto what you need.
YUCK!
Quoted 9 lines
Quoted 2 lines
So using the uimage format is much more
flexible.I agree.
So I'd see the following changes for this patch:
- Reword "fix" to show enhancement, and detail the flexibility rationale
in the description.
- Extend to u-boot-with-spl.imx.
Ok.
Tom, shall we merge this one into current release after it's fixed?
PATCH 3/6 arm: imx: Fix u-boot-with-nand-spl.imx target
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Quoted 40 lines · Marek Vasut
Dear Beno?t Th?baudeau,
Quoted 38 lines
Dear Marek Vasut,
Quoted 35 lines · Marek Vasut
Dear Beno?t Th?baudeau,
Quoted 33 lines
Dear Marek Vasut,
Quoted 6 lines · Marek Vasut
This target is currently concatenating u-boot SPL in imximage format with u-boot.bin. The NAND SPL can load a raw binary, but the preffered format with much less limitations is uImage format. Fix the target so u-boot.bin is first converted into uImage format and only after that is concatenated.
Please reword: This is not a fix, but an enhancement.
Quoted 21 lines
Signed-off-by Marek Vasut <[email protected]> Cc: AlbertARIBAUD <[email protected]> Cc: Beno?t Th?baudeau <[email protected]> Cc: Fabio Estevam <[email protected]> Cc: Scott Wood <[email protected]> Cc: Stefano Babic <[email protected]> Cc: Tom Rini <[email protected]> ---
arch/arm/imx-common/Makefile| 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)(quoted diff) +0
-0b/arch/arm/imx-common/Makefile index 44b6822..ba31d3e 100644 --- a/arch/arm/imx-common/Makefile +++ b/arch/arm/imx-common/Makefile @@ -69,8 +69,11 @@ $(OBJTREE)/u-boot-with-nand-spl.imx: $(OBJTREE)/SPL $(OBJTREE)/u-boot.bin
-I binary -O binary $(OBJTREE)/spl/u-boot-nand-spl.imx \
$(OBJTREE)/spl/u-boot-nand-spl-pad.imx rm $(OBJTREE)/spl/u-boot-nand-spl.imx
Quoted 18 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
(quoted diff) +0
-1- cat $(OBJTREE)/spl/u-boot-nand-spl-pad.imx$(OBJTREE)/u-boot.bin
$@ - rm $(OBJTREE)/spl/u-boot-nand-spl-pad.imx +
$(OBJTREE)/tools/mkimage -A arm -O U-Boot -a
$(CONFIG_SYS_TEXT_BASE) \ + -e $(CONFIG_SYS_TEXT_BASE) -Cnone -d
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
$(OBJTREE)/u-boot.bin \ + $(OBJTREE)/u-boot.uim + cat
$(OBJTREE)/spl/u-boot-nand-spl-pad.imx
$(OBJTREE)/u-boot.uim$@ + rm $(OBJTREE)/spl/u-boot-nand-spl-pad.imx
$(OBJTREE)/u-boot.uim####################################################################
#### #
Quoted 26 lines
Quoted 12 lines
Quoted 12 lines
Quoted 11 lines
Quoted 2 lines
-- 1.7.11.7
This change is correct.
I don't know the uImage format very well. Can you give examples of what this patch can bring compared to the raw binary format?
If this change is useful, it should probably be made for u-boot-with-spl.imx too.
Quoted 7 lines
Quoted 5 lines
Please see
common/spl/spl_nand.cwhich callsspl_parse_image_header()on the payload loaded from NAND. This is implemented incommon/spl/spl.cand if proper header is not found, the payload is assumed to be of 200kB of size, which is far less than fully configured U-Boot.Or you could #define
CONFIG_SYS_MONITOR_LENto what you need.YUCK!
Quoted 4 lines
Quoted 1 line
So using the uimage format is much more flexible.
I agree.
Quoted 8 lines
Quoted 3 lines
So I'd see the following changes for this patch: - Reword "fix" to show enhancement, and detail the flexibility rationale in the description. - Extend to u-boot-with-spl.imx.
Ok.
Tom, shall we merge this one into current release after it's fixed?
This is an enhancement not a bug fix, right? If so, next release.
- --
Tom
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
iQIcBAEBAgAGBQJRcTCWAAoJENk4IS6UOR1WFx8P/1CXwEcPqrvZthAJPqMYOjvU r1FVOX0MgXCkRQ1FXnGKaYdlJbc+1GS09KgsnbQ+AOwMEPPPAju62npHzP1/CfhH FkRInGmb/cqSuIt3shXg/Q4URkGEyMW6QhQpaskMgLNAzjGX1Q12Wnz2O1BLnIzr 1y9hMxNaqA8aIW9apXEtC/U67vHFVU74jT5IHzv/F6rzcAayoCz2Az0oG4bvRR90 HTPJMLH/z2lis1JgG5GSiH+604z+KOprNz1Uid2aZr7os/WmDsPkO4rJ0l8kd2HE Us7Au7/b0h1TfXrJv6oYNEsYn4jECPZ8HeHxjcRTi102JBAiiAK914PSuKVm+Kys ILffCZXcOHS6Q3zUTp4EkrcrxVzJxX+yIv2SJipy0dXx7mmrM559J55DUyFLsrKh o8Z4MMidjqulmTdQNeBhLLIo80URynouRQ2ZpiwpR7iKdObFuAPr0HVzg65tNpbo vr5WOquk4fooK88s3HXKjDvIyBJcq08rtn5sgmJCvba57WRrlnYwS1D3SF6NT/Me isF6jeL7nsZF2mckeCHJ3Wduw6kbg9n9mbfDWWqpYyqh/srvWvN6A7gCtwrlK2eR vN8DvDkBpWWulqfRzq6PLs8nH1x3WOajiA9iICc1O7+YyXzKGF+sAEBcBgTcH3fX lz4Hs0VAu3RA/kdeJiVJ
=zMqB
-----END PGP SIGNATURE-----
PATCH 3/6 arm: imx: Fix u-boot-with-nand-spl.imx target
Dear Tom Rini,
Quoted 90 lines · Marek Vasut
Quoted 40 lines
Dear Beno?t Th?baudeau,
Quoted 38 lines
Dear Marek Vasut,
Quoted 35 lines · Marek Vasut
Dear Beno?t Th?baudeau,
Quoted 33 lines
Dear Marek Vasut,
Quoted 6 lines · Marek Vasut
This target is currently concatenating u-boot SPL in imximage format with u-boot.bin. The NAND SPL can load a raw binary, but the preffered format with much less limitations is uImage format. Fix the target so u-boot.bin is first converted into uImage format and only after that is concatenated.
Please reword: This is not a fix, but an enhancement.
Quoted 21 lines
Signed-off-by Marek Vasut <[email protected]> Cc: AlbertARIBAUD <[email protected]> Cc: Beno?t Th?baudeau <[email protected]> Cc: Fabio Estevam <[email protected]> Cc: Scott Wood <[email protected]> Cc: Stefano Babic <[email protected]> Cc: Tom Rini <[email protected]> ---
arch/arm/imx-common/Makefile| 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)(quoted diff) +0
-0b/arch/arm/imx-common/Makefile index 44b6822..ba31d3e 100644 --- a/arch/arm/imx-common/Makefile +++ b/arch/arm/imx-common/Makefile @@ -69,8 +69,11 @@ $(OBJTREE)/u-boot-with-nand-spl.imx: $(OBJTREE)/SPL $(OBJTREE)/u-boot.bin
-I binary -O binary $(OBJTREE)/spl/u-boot-nand-spl.imx \
$(OBJTREE)/spl/u-boot-nand-spl-pad.imx rm $(OBJTREE)/spl/u-boot-nand-spl.imxQuoted 17 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
(quoted diff) +0
-1- cat $(OBJTREE)/spl/u-boot-nand-spl-pad.imx$(OBJTREE)/u-boot.bin
$@ - rm $(OBJTREE)/spl/u-boot-nand-spl-pad.imx +
$(OBJTREE)/tools/mkimage -A arm -O U-Boot -a
$(CONFIG_SYS_TEXT_BASE) \ + -e $(CONFIG_SYS_TEXT_BASE) -Cnone -d
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
$(OBJTREE)/u-boot.bin \ + $(OBJTREE)/u-boot.uim + cat
$(OBJTREE)/spl/u-boot-nand-spl-pad.imx
$(OBJTREE)/u-boot.uim$@ + rm $(OBJTREE)/spl/u-boot-nand-spl-pad.imx
$(OBJTREE)/u-boot.uim####################################################################
#### #
Quoted 29 lines
Quoted 11 lines
Quoted 11 lines
Quoted 10 lines
Quoted 1 line
-- 1.7.11.7
This change is correct.
I don't know the uImage format very well. Can you give examples of what this patch can bring compared to the raw binary format?
If this change is useful, it should probably be made for u-boot-with-spl.imx too.
Quoted 7 lines
Quoted 5 lines
Please see
common/spl/spl_nand.cwhich callsspl_parse_image_header()on the payload loaded from NAND. This is implemented incommon/spl/spl.cand if proper header is not found, the payload is assumed to be of 200kB of size, which is far less than fully configured U-Boot.Or you could #define
CONFIG_SYS_MONITOR_LENto what you need.YUCK!
Quoted 7 lines
Quoted 1 line
So using the uimage format is much more flexible.
I agree.
So I'd see the following changes for this patch: - Reword "fix" to show enhancement, and detail the flexibility rationale in the description. - Extend to u-boot-with-spl.imx.
Quoted 6 lines
Quoted 4 lines
Ok.
Tom, shall we merge this one into current release after it's fixed?
This is an enhancement not a bug fix, right? If so, next release.
Since the release is nigh, let's wrap it into the next one. But the usability of the spl is lower :(
PATCH 4/6 arm: mx5: Add SPL support code to MX5
In reply to Marek VasutFix minor adjustments needed to get SPL framework building on MX5.
--- arch/arm/include/asm/arch-mx5/spl.h | 19 +++++++++++++++++++ spl/Makefile | 4 ++++ 2 files changed, 23 insertions(+) create mode 100644 arch/arm/include/asm/arch-mx5/spl.h
arch/arm/include/asm/arch-mx5/spl.h new file +19 -0
@@ -0,0 +1,19 @@ +/* +* Copyright (C) 2013 Marek Vasut <[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. +*/ + +#ifndef __ASM_ARCH_SPL_H__ +#define __ASM_ARCH_SPL_H__ + +#define BOOT_DEVICE_NONE 0 +#define BOOT_DEVICE_NAND 1 + +#endif /* __ASM_ARCH_SPL_H__ */
spl/Makefile +4 -0
@@ -88,6 +88,10 @@ ifneq ($(CONFIG_AM33XX)$(CONFIG_OMAP34XX)$(CONFIG_OMAP44XX)$(CONFIG_OMAP54XX)$(C LIBS-y += $(CPUDIR)/omap-common/libomap-common.o endif +ifneq (,$(filter $(SOC), mx25 mx27 mx5 mx6 mx31 mx35)) +LIBS-y += arch/$(ARCH)/imx-common/libimx-common.o +endif + ifneq ($(CONFIG_TEGRA),) LIBS-y += arch/$(ARCH)/cpu/$(SOC)-common/lib$(SOC)-common.o LIBS-y += arch/$(ARCH)/cpu/tegra-common/libcputegra-common.o
Signature · 1 line
1.7.11.7
PATCH 4/6 arm: mx5: Add SPL support code to MX5
Dear Marek Vasut,
Quoted 58 lines · Marek Vasut
Fix minor adjustments needed to get SPL framework building on MX5.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>--- arch/arm/include/asm/arch-mx5/spl.h | 19 +++++++++++++++++++ spl/Makefile | 4 ++++ 2 files changed, 23 insertions(+) create mode 100644 arch/arm/include/asm/arch-mx5/spl.harch/arm/include/asm/arch-mx5/spl.h new file +19
-0@@ -0,0 +1,19 @@ +/* +* Copyright (C) 2013 Marek Vasut <[email protected]> +* +* See file CREDITS for list of people who contributed to this +* project. +* +* This program is free software; you can redistribute it and/or +* modify it under the terms of the GNU General Public License as +* published by the Free Software Foundation; either version 2 of +* the License, or (at your option) any later version. +*/ + +#ifndef __ASM_ARCH_SPL_H__ +#define __ASM_ARCH_SPL_H__ + +#define BOOT_DEVICE_NONE 0 +#define BOOT_DEVICE_NAND 1 + +#endif /* __ASM_ARCH_SPL_H__ */ spl/Makefile +0
-0@@ -88,6 +88,10 @@ ifneq($(CONFIG_AM33XX)$(CONFIG_OMAP34XX)$(CONFIG_OMAP44XX)$(CONFIG_OMAP54XX)$(C
LIBS-y += $(CPUDIR)/omap-common/libomap-common.o
endifspl/Makefile +4
-1+ifneq (,$(filter $(SOC), mx25 mx27 mx5 mx6 mx31 mx35)) +LIBS-y += arch/$(ARCH)/imx-common/libimx-common.o +endif + ifneq ($(CONFIG_TEGRA),) LIBS-y += arch/$(ARCH)/cpu/$(SOC)-common/lib$(SOC)-common.o LIBS-y += arch/$(ARCH)/cpu/tegra-common/libcputegra-common.o --
Quoted 1 line
1.7.11.7
Best regards,
Beno?t
PATCH 4/6 arm: mx5: Add SPL support code to MX5
In reply to Marek VasutQuoted 1 line · Marek Vasut
Fix minor adjustments needed to get SPL framework building on MX5.
The same adjustment is needed to make other mx* build. Perhaps spl.h should live under arch/arm/include/asm/imx-common to avoid duplicating it for others?
- Philip
Signature · 3 lines
Philip Paeps Senior Reality Engineer Ministry of Information
PATCH 4/6 arm: mx5: Add SPL support code to MX5
Dear Philip Paeps,
Quoted 6 lines · Marek Vasut
Quoted 1 line
Fix minor adjustments needed to get SPL framework building on MX5.
The same adjustment is needed to make other mx* build. Perhaps
spl.hshould live underarch/arm/include/asm/imx-commonto avoid duplicating it for others?
Yes, sounds reasonable.
Quoted 6 lines
- Philip
Signature · 3 lines
Philip Paeps Senior Reality Engineer Ministry of Information
Is that from 1984 ? :-)
Best regards,
Marek Vasut
PATCH 4/6 arm: mx5: Add SPL support code to MX5
In reply to Philip PaepsDear Philip Paeps,
Quoted 6 lines · Marek Vasut
Quoted 1 line
Fix minor adjustments needed to get SPL framework building on MX5.
The same adjustment is needed to make other mx* build. Perhaps
spl.hshould live underarch/arm/include/asm/imx-commonto avoid duplicating it for others?
Apparently, this doesn't work. The include isn't found.
Best regards,
Marek Vasut
PATCH 4/6 arm: mx5: Add SPL support code to MX5
In reply to Marek VasutQuoted 10 lines · Marek Vasut
Fix minor adjustments needed to get SPL framework building on MX5.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>---
Applied to u-boot-imx, thanks.
Best regards,
Stefano Babic
Signature · 5 lines
===================================================================== DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: +49-8142-66989-53 Fax: +49-8142-66989-80 Email: sbabic at denx.de =====================================================================
PATCH 5/6 arm: mx5: Add NAND clock handling
In reply to Marek VasutAugment the MX5 clock code with function to enable and configure NFC clock. This is necessary to get NFC working on MX5.
--- arch/arm/cpu/armv7/mx5/clock.c | 14 ++++++++++++-- arch/arm/include/asm/arch-mx5/clock.h | 1 + 2 files changed, 13 insertions(+), 2 deletions(-)
arch/arm/cpu/armv7/mx5/clock.c +12 -2
@@ -739,10 +739,11 @@ static int config_core_clk(u32 ref, u32 freq)static int config_nfc_clk (u32 nfc_clk) {u32 parent_rate =get_emi_slow_clk (); -u32 div = parent_rate / nfc_clk; +u32 div; -if (nfc_clk <=0 ) +if (nfc_clk ==0 )return -EINVAL ; + div = parent_rate / nfc_clk;if (div ==0 ) div++;if (parent_rate / div >NFC_CLK_MAX )
@@ -755,6 +756,15 @@ static int config_nfc_clk(u32 nfc_clk)return 0 ; } +void enable_nfc_clk (unsigned char enable) +{ +unsigned int cg = enable ?MXC_CCM_CCGR_CG_ON :MXC_CCM_CCGR_CG_OFF ; + +clrsetbits_le32 (&mxc_ccm->CCGR5, +MXC_CCM_CCGR5_EMI_ENFC (MXC_CCM_CCGR_CG_MASK ), +MXC_CCM_CCGR5_EMI_ENFC (cg)); +} +/* Config main_bus_clock for periphs */ static int config_periph_clk (u32 ref,u32 freq) {
arch/arm/include/asm/arch-mx5/clock.h +1 -0
@@ -68,5 +68,6 @@ void set_usboh3_clk(void);void enable_usboh3_clk (unsigned char enable);void mxc_set_sata_internal_clock (void );int enable_i2c_clk (unsigned char enable,unsigned i2c_num); +void enable_nfc_clk (unsigned char enable);#endif /* __ASM_ARCH_CLOCK_H */
Signature · 1 line
1.7.11.7
PATCH 5/6 arm: mx5: Add NAND clock handling
Dear Marek Vasut,
Quoted 62 lines · Marek Vasut
Augment the MX5 clock code with function to enable and configure NFC clock. This is necessary to get NFC working on MX5.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>--- arch/arm/cpu/armv7/mx5/clock.c | 14 ++++++++++++-- arch/arm/include/asm/arch-mx5/clock.h | 1 + 2 files changed, 13 insertions(+), 2 deletions(-)arch/arm/cpu/armv7/mx5/clock.c +12
-2@@ -739,10 +739,11 @@ static int config_core_clk(u32 ref, u32 freq)static int config_nfc_clk (u32 nfc_clk) {u32 parent_rate =get_emi_slow_clk (); -u32 div = parent_rate / nfc_clk; +u32 div; -if (nfc_clk <=0 ) +if (nfc_clk ==0 )return -EINVAL ; + div = parent_rate / nfc_clk;if (div ==0 ) div++;if (parent_rate / div >NFC_CLK_MAX )@@ -755,6 +756,15 @@ static int config_nfc_clk(u32 nfc_clk)return 0 ; } +void enable_nfc_clk (unsigned char enable) +{ +unsigned int cg = enable ?MXC_CCM_CCGR_CG_ON :MXC_CCM_CCGR_CG_OFF ; + +clrsetbits_le32 (&mxc_ccm->CCGR5, +MXC_CCM_CCGR5_EMI_ENFC (MXC_CCM_CCGR_CG_MASK ), +MXC_CCM_CCGR5_EMI_ENFC (cg)); +} +/* Config main_bus_clock for periphs */ static int config_periph_clk (u32 ref,u32 freq) {arch/arm/include/asm/arch-mx5/clock.h +1
-1@@ -68,5 +68,6 @@ void set_usboh3_clk(void);void enable_usboh3_clk (unsigned char enable);void mxc_set_sata_internal_clock (void );int enable_i2c_clk (unsigned char enable,unsigned i2c_num); +void enable_nfc_clk (unsigned char enable);#endif /* __ASM_ARCH_CLOCK_H */ --
Quoted 1 line
1.7.11.7
Best regards,
Beno?t
PATCH 5/6 arm: mx5: Add NAND clock handling
Quoted 64 lines · Beno?t Th?baudeau
Dear Marek Vasut,
Quoted 61 lines · Marek Vasut
Augment the MX5 clock code with function to enable and configure NFC clock. This is necessary to get NFC working on MX5.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>--- arch/arm/cpu/armv7/mx5/clock.c | 14 ++++++++++++-- arch/arm/include/asm/arch-mx5/clock.h | 1 + 2 files changed, 13 insertions(+), 2 deletions(-)arch/arm/cpu/armv7/mx5/clock.c +12
-2@@ -739,10 +739,11 @@ static int config_core_clk(u32 ref, u32 freq)static int config_nfc_clk (u32 nfc_clk) {u32 parent_rate =get_emi_slow_clk (); -u32 div = parent_rate / nfc_clk; +u32 div; -if (nfc_clk <=0 ) +if (nfc_clk ==0 )return -EINVAL ; + div = parent_rate / nfc_clk;if (div ==0 ) div++;if (parent_rate / div >NFC_CLK_MAX )@@ -755,6 +756,15 @@ static int config_nfc_clk(u32 nfc_clk)return 0 ; } +void enable_nfc_clk (unsigned char enable) +{ +unsigned int cg = enable ?MXC_CCM_CCGR_CG_ON :MXC_CCM_CCGR_CG_OFF ; + +clrsetbits_le32 (&mxc_ccm->CCGR5, +MXC_CCM_CCGR5_EMI_ENFC (MXC_CCM_CCGR_CG_MASK ), +MXC_CCM_CCGR5_EMI_ENFC (cg)); +} +/* Config main_bus_clock for periphs */ static int config_periph_clk (u32 ref,u32 freq) {arch/arm/include/asm/arch-mx5/clock.h +1
-0@@ -68,5 +68,6 @@ void set_usboh3_clk(void);void enable_usboh3_clk (unsigned char enable);void mxc_set_sata_internal_clock (void );int enable_i2c_clk (unsigned char enable,unsigned i2c_num); +void enable_nfc_clk (unsigned char enable);
Quoted 5 lines
Quoted 3 lines
arch/arm/include/asm/arch-mx5/clock.h +0
-1#endif /* __ASM_ARCH_CLOCK_H */ --1.7.11.7
Reviewed-by Beno?t Th?baudeau <[email protected]>
Actually, this code is correct, but why is it needed? Indeed, this clock is already ungated by arch/arm/cpu/armv7/mx5/lowlevel_init.S. Also, your board seems to be hardware-bootable only from NAND, so in this case the boot ROM also ungates this clock itself.
Best regards,
Beno?t
PATCH 5/6 arm: mx5: Add NAND clock handling
Dear Beno?t Th?baudeau,
Quoted 83 lines · Beno?t Th?baudeau
Quoted 82 lines
Dear Marek Vasut,
Quoted 79 lines · Marek Vasut
Augment the MX5 clock code with function to enable and configure NFC clock. This is necessary to get NFC working on MX5.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>---
arch/arm/cpu/armv7/mx5/clock.c | 14 ++++++++++++-- arch/arm/include/asm/arch-mx5/clock.h | 1 + 2 files changed, 13 insertions(+), 2 deletions(-)arch/arm/cpu/armv7/mx5/clock.c +12
-2@@ -739,10 +739,11 @@ static int config_core_clk(u32 ref, u32 freq)static int config_nfc_clk (u32 nfc_clk) {u32 parent_rate =get_emi_slow_clk (); -u32 div = parent_rate / nfc_clk; +u32 div; -if (nfc_clk <=0 ) +if (nfc_clk ==0 )return -EINVAL ; + div = parent_rate / nfc_clk;if (div ==0 ) div++;if (parent_rate / div >NFC_CLK_MAX )@@ -755,6 +756,15 @@ static int config_nfc_clk(u32 nfc_clk)return 0 ; } +void enable_nfc_clk (unsigned char enable) +{ +unsigned int cg = enable ?MXC_CCM_CCGR_CG_ON :MXC_CCM_CCGR_CG_OFF ; + +clrsetbits_le32 (&mxc_ccm->CCGR5, +MXC_CCM_CCGR5_EMI_ENFC (MXC_CCM_CCGR_CG_MASK ), +MXC_CCM_CCGR5_EMI_ENFC (cg)); +} +/* Config
main_bus_clockfor periphs */
static intconfig_periph_clk(u32 ref, u32 freq)
{arch/arm/include/asm/arch-mx5/clock.h +1
-0@@ -68,5 +68,6 @@ void set_usboh3_clk(void);void enable_usboh3_clk (unsigned char enable);void mxc_set_sata_internal_clock (void );int enable_i2c_clk (unsigned char enable,unsigned i2c_num); +void enable_nfc_clk (unsigned char enable);#endif /*
__ASM_ARCH_CLOCK_H*/
Quoted 7 lines
Quoted 4 lines
Quoted 2 lines
--
1.7.11.7Reviewed-by Beno?t Th?baudeau <[email protected]>Actually, this code is correct, but why is it needed? Indeed, this clock is already ungated by
arch/arm/cpu/armv7/mx5/lowlevel_init.S.
This is not good, the code should be moved away from the assembly.
Quoted 2 lines
Also, your board
seems to be hardware-bootable only from NAND
This is incorrect, it is bootable from SD card as well.
Quoted 2 lines
so in this case the boot ROM
also ungates this clock itself.
We can not rely on that.
PATCH 5/6 arm: mx5: Add NAND clock handling
Dear Marek Vasut,
Quoted 98 lines · Marek Vasut
Dear Beno?t Th?baudeau,
Quoted 83 lines · Beno?t Th?baudeau
Quoted 82 lines
Dear Marek Vasut,
Quoted 79 lines · Marek Vasut
Augment the MX5 clock code with function to enable and configure NFC clock. This is necessary to get NFC working on MX5.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>---
arch/arm/cpu/armv7/mx5/clock.c | 14 ++++++++++++-- arch/arm/include/asm/arch-mx5/clock.h | 1 + 2 files changed, 13 insertions(+), 2 deletions(-)arch/arm/cpu/armv7/mx5/clock.c +12
-2@@ -739,10 +739,11 @@ static int config_core_clk(u32 ref, u32 freq)static int config_nfc_clk (u32 nfc_clk) {u32 parent_rate =get_emi_slow_clk (); -u32 div = parent_rate / nfc_clk; +u32 div; -if (nfc_clk <=0 ) +if (nfc_clk ==0 )return -EINVAL ; + div = parent_rate / nfc_clk;if (div ==0 ) div++;if (parent_rate / div >NFC_CLK_MAX )@@ -755,6 +756,15 @@ static int config_nfc_clk(u32 nfc_clk)return 0 ; } +void enable_nfc_clk (unsigned char enable) +{ +unsigned int cg = enable ?MXC_CCM_CCGR_CG_ON :MXC_CCM_CCGR_CG_OFF ; + +clrsetbits_le32 (&mxc_ccm->CCGR5, +MXC_CCM_CCGR5_EMI_ENFC (MXC_CCM_CCGR_CG_MASK ), +MXC_CCM_CCGR5_EMI_ENFC (cg)); +} +/* Config
main_bus_clockfor periphs */
static intconfig_periph_clk(u32 ref, u32 freq)
{arch/arm/include/asm/arch-mx5/clock.h +1
-0@@ -68,5 +68,6 @@ void set_usboh3_clk(void);void enable_usboh3_clk (unsigned char enable);void mxc_set_sata_internal_clock (void );int enable_i2c_clk (unsigned char enable,unsigned i2c_num); +void enable_nfc_clk (unsigned char enable);#endif /*
__ASM_ARCH_CLOCK_H*/Quoted 7 lines
Quoted 4 lines
Quoted 2 lines
--
1.7.11.7Reviewed-by Beno?t Th?baudeau <[email protected]>Actually, this code is correct, but why is it needed? Indeed, this clock is already ungated by
arch/arm/cpu/armv7/mx5/lowlevel_init.S.This is not good, the code should be moved away from the assembly.
Quoted 2 lines
Also, your board
seems to be hardware-bootable only from NAND
Quoted 6 lines
This is incorrect, it is bootable from SD card as well.
Quoted 2 lines
so in this case the boot ROM
also ungates this clock itself.We can not rely on that.
OK, then this patch is fine.
Best regards,
Beno?t
PATCH 5/6 arm: mx5: Add NAND clock handling
In reply to Marek VasutQuoted 11 lines · Marek Vasut
Augment the MX5 clock code with function to enable and configure NFC clock. This is necessary to get NFC working on MX5.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>---
Applied to u-boot-imx, thanks.
Best regards,
Stefano Babic
Signature · 5 lines
===================================================================== DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: +49-8142-66989-53 Fax: +49-8142-66989-80 Email: sbabic at denx.de =====================================================================
PATCH 6/6 arm: mx5: Add support for DENX M53EVK
In reply to Marek VasutAdd basic support for the DENX M53EVK board. Currently supported is the MMC, Ethernet, I2C.
--- MAINTAINERS | 1 + board/denx/m53evk/Makefile | 40 ++++ board/denx/m53evk/imximage.cfg | 108 +++++++++++ board/denx/m53evk/m53evk.c | 410 +++++++++++++++++++++++++++++++++++++++++ boards.cfg | 1 + include/configs/m53evk.h | 252 +++++++++++++++++++++++++ 6 files changed, 812 insertions(+) create mode 100644 board/denx/m53evk/Makefile create mode 100644 board/denx/m53evk/imximage.cfg create mode 100644 board/denx/m53evk/m53evk.c create mode 100644 include/configs/m53evk.h
MAINTAINERS +1 -0
@@ -954,6 +954,7 @@ Marek Vasut <[email protected]> mx23_olinuxino i.MX23 m28evk i.MX28 sc_sps_1 i.MX28 + m53evk i.MX53 Hugo Villeneuve <[email protected]>
board/denx/m53evk/Makefile new file +40 -0
@@ -0,0 +1,40 @@ +# +# DENX M53EVK +# Copyright (C) 2012-2013 Marek Vasut <[email protected]> +# +# 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 $(TOPDIR)/config.mk + +LIB = $(obj)lib$(BOARD).o + +COBJS := m53evk.o + +SRCS := $(COBJS:.o=.c) +OBJS := $(addprefix $(obj),$(COBJS)) + +$(LIB): $(obj).depend $(OBJS) + $(call cmd_link_o_target, $(OBJS)) + +######################################################################### + +# defines $(obj).depend target +include $(SRCTREE)/rules.mk + +sinclude $(obj).depend + +#########################################################################
board/denx/m53evk/imximage.cfg new file +108 -0
108 lines not on this page · raw message
board/denx/m53evk/m53evk.c new file +410 -0
410 lines not on this page · raw message
boards.cfg +1 -0
@@ -246,6 +246,7 @@ am335x_evm_usbspl arm armv7 am335x ti ti814x_evm arm armv7 ti814x ti am33xx pcm051 arm armv7 pcm051 phytec am33xx pcm051 highbank arm armv7 highbank - highbank +m53evk arm armv7 m53evk denx mx5 m53evk:IMX_CONFIG=board/denx/m53evk/imximage.cfg mx51_efikamx arm armv7 mx51_efikamx genesi mx5 mx51_efikamx:MACH_TYPE=MACH_TYPE_MX51_EFIKAMX,IMX_CONFIG=board/genesi/mx51_efikamx/imximage_mx.cfg mx51_efikasb arm armv7 mx51_efikamx genesi mx5 mx51_efikamx:MACH_TYPE=MACH_TYPE_MX51_EFIKASB,IMX_CONFIG=board/genesi/mx51_efikamx/imximage_sb.cfg mx51evk arm armv7 mx51evk freescale mx5 mx51evk:IMX_CONFIG=board/freescale/mx51evk/imximage.cfg
include/configs/m53evk.h new file +252 -0
252 lines not on this page · raw message
Signature · 1 line
1.7.11.7
PATCH 6/6 arm: mx5: Add support for DENX M53EVK
Dear Marek Vasut,
In message <[email protected]> you wrote:
...
include/configs/m53evk.h line 45
Quoted 4 lines
include/configs/m53evk.h new file +4
-0+#define CONFIG_CMD_DHCP +#define CONFIG_CMD_EXT2 +#define CONFIG_CMD_FAT +#define CONFIG_CMD_FAT
One CONFIG_CMD_FAT should be enough.
include/configs/m53evk.h line 53
Quoted 8 lines
include/configs/m53evk.h new file +8
-0+#define CONFIG_CMD_I2C +#define CONFIG_CMD_MII +#define CONFIG_CMD_MMC +#define CONFIG_CMD_NAND +#define CONFIG_CMD_NET +#define CONFIG_CMD_PING +#define CONFIG_CMD_SATA +#define CONFIG_CMD_USB
As CONFIG_CMD_DATE is not set, we should enable CONFIG_TIMESTAMP.
include/configs/m53evk.h line 164
Quoted 6 lines
include/configs/m53evk.h new file +6
-0+/* +* Ethernet on SOC (FEC) +*/ +#ifdef CONFIG_CMD_NET +#define CONFIG_FEC_MXC +#define CONFIG_ETHPRIME "FEC0"
What would that be good for? We have only a single network interface, so please drop that.
Quoted 1 line
+#define
CONFIG_ARP_TIMEOUT200UL
Is this really needed?
include/configs/m53evk.h line 217
Quoted 3 lines
include/configs/m53evk.h new file +3
-0+#define CONFIG_CMDLINE_TAG +#define CONFIG_INITRD_TAG +#define CONFIG_SETUP_MEMORY_TAGS
I think we support only DT enabled kernels, so do we really need these?
Quoted 1 line
+#define
CONFIG_BOOTFILE"uImage"
Please make this "m53evk/uImage" as usual.
Don't we need a DT file as well?
include/configs/m53evk.h line 221
Quoted 2 lines
include/configs/m53evk.h new file +2
-0+#define CONFIG_BOOTARGS "console=ttymxc1,115200" +#define CONFIG_BOOTCOMMAND "run bootcmd_net"
Where is "bootcmd_net" defined?
I doubt that this is actually a working environment.
include/configs/m53evk.h line 223
Quoted 2 lines
include/configs/m53evk.h new file +2
-0+#define CONFIG_LOADADDR 0x70800000 +#define CONFIG_SYS_LOAD_ADDR CONFIG_LOADADDR
What do we need CONFIG_LOADADDR for?
include/configs/m53evk.h line 242
Quoted 11 lines
include/configs/m53evk.h new file +11
-0+/* +* The 0x1000 offset must be present, otherwise board won't boot. +* The value (IRAM_BASE_ADDR) must be spelled out, otherwise linker won't link. +*/ +#define CONFIG_SPL_TEXT_BASE 0x70008000 +#define CONFIG_SPL_PAD_TO 0x8000 +#define CONFIG_SPL_STACK 0x70004000 +#define CONFIG_SPL_LIBCOMMON_SUPPORT +#define CONFIG_SPL_LIBGENERIC_SUPPORT +#define CONFIG_SPL_SERIAL_SUPPORT +#define CONFIG_SPL_GPIO_SUPPORT
The comment above does not relate to the definitions here. Is it misplaced, or not correct?
Best regards,
Wolfgang Denk
Signature · 5 lines
DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de "Just Say No." - Nancy Reagan "No." - Ronald Reagan
PATCH 6/6 arm: mx5: Add support for DENX M53EVK
Dear Wolfgang Denk,
Quoted 11 lines
Dear Marek Vasut,
In message <[email protected]> you wrote:
...include/configs/m53evk.h line 45
Quoted 4 lines
include/configs/m53evk.h new file +4
-0+#define CONFIG_CMD_DHCP +#define CONFIG_CMD_EXT2 +#define CONFIG_CMD_FAT +#define CONFIG_CMD_FAT One
CONFIG_CMD_FATshould be enough.
Indeed, it might be too fat now.
[...]
Quoted 7 lines
include/configs/m53evk.h line 217
Quoted 4 lines
include/configs/m53evk.h new file +3
-0+#define CONFIG_CMDLINE_TAG +#define CONFIG_INITRD_TAG +#define CONFIG_SETUP_MEMORY_TAGS I think we support only DT enabled kernels, so do we really need these?
I do need those to boot ancient FSL kernel (for that Android 4.1.2 for MX53 with working graphics acceleration I'm cooking here).
Quoted 5 lines
Quoted 1 line
+#define
CONFIG_BOOTFILE"uImage"Please make this "m53evk/uImage" as usual.
Don't we need a DT file as well?
OK
Quoted 6 lines
include/configs/m53evk.h line 221
Quoted 2 lines
include/configs/m53evk.h new file +2
-0+#define CONFIG_BOOTARGS "console=ttymxc1,115200" +#define CONFIG_BOOTCOMMAND "run bootcmd_net"Where is "
bootcmd_net" defined?I doubt that this is actually a working environment.
Right.
Quoted 4 lines
include/configs/m53evk.h line 223
Quoted 2 lines
include/configs/m53evk.h new file +2
-0+#define CONFIG_LOADADDR 0x70800000 +#define CONFIG_SYS_LOAD_ADDR CONFIG_LOADADDRWhat do we need
CONFIG_LOADADDRfor?
For this stuff, the "loadaddr" env variable. I have kinda idea to kill this and unify it to CONFIG_SYS_LOAD_ADDR all around, what do you think?
include/env_default.h:#ifdef CONFIG_LOADADDRinclude/env_default.h: "loadaddr=" __stringify(CONFIG_LOADADDR) "\0"
Quoted 6 lines
Quoted 6 lines
+/*
+ * The0x1000offset must be present, otherwise board won't boot.
+ * The value (IRAM_BASE_ADDR) must be spelled out, otherwise linker won't link. + */include/configs/m53evk.h new file +2
-0+#define CONFIG_SPL_TEXT_BASE 0x70008000 +#define CONFIG_SPL_PAD_TO 0x8000
Quoted 8 lines
include/configs/m53evk.h line 242
Quoted 5 lines
include/configs/m53evk.h new file +5
-0+#define CONFIG_SPL_STACK 0x70004000 +#define CONFIG_SPL_LIBCOMMON_SUPPORT +#define CONFIG_SPL_LIBGENERIC_SUPPORT +#define CONFIG_SPL_SERIAL_SUPPORT +#define CONFIG_SPL_GPIO_SUPPORT The comment above does not relate to the definitions here. Is it misplaced, or not correct?
Damn, remnant of my experimentation.
PATCH 6/6 arm: mx5: Add support for DENX M53EVK
Dear Marek Vasut,
Quoted 8 lines · In message <[email protected]> you wrote:
Quoted 5 lines
include/configs/m53evk.h line 223
Quoted 3 lines
include/configs/m53evk.h new file +2
-0+#define CONFIG_LOADADDR 0x70800000 +#define CONFIG_SYS_LOAD_ADDR CONFIG_LOADADDRWhat do we need
CONFIG_LOADADDRfor?For this stuff, the "loadaddr" env variable. I have kinda idea to kill this and unify it to
CONFIG_SYS_LOAD_ADDRall around, what do you think?
Why CONFIG_SYS_* ?
Quoted 2 lines
include/env_default.h:#ifdef
CONFIG_LOADADDRinclude/env_default.h: "loadaddr="__stringify(CONFIG_LOADADDR)"\0"
Indeed - I missed this one. I only checked C and asm sources, and cound not find a single reference...
So arch/blackfin/include/asm/config.h and include/env_default.h are the only locateions that refer to CONFIG_LOADADDR ? This should indeed be cleaned up. [But this is not related to your patch any longer.]
Best regards,
Wolfgang Denk
Signature · 5 lines
DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de If something is different, it's either better or worse, and usually both. - Larry Wall
PATCH 6/6 arm: mx5: Add support for DENX M53EVK
Dear Wolfgang Denk,
Quoted 8 lines
Dear Marek Vasut,
Quoted 5 lines · In message <[email protected]> you wrote:
Quoted 4 lines
include/configs/m53evk.h line 223
Quoted 2 lines
include/configs/m53evk.h new file +2
-0+#define CONFIG_LOADADDR 0x70800000 +#define CONFIG_SYS_LOAD_ADDR CONFIG_LOADADDRWhat do we need
CONFIG_LOADADDRfor?
Quoted 4 lines
Quoted 2 lines
For this stuff, the "loadaddr" env variable. I have kinda idea to kill this and unify it to
CONFIG_SYS_LOAD_ADDRall around, what do you think?Why CONFIG_SYS_* ?
What do you suggest then?
Quoted 6 lines
Quoted 3 lines
include/env_default.h:#ifdef
CONFIG_LOADADDRinclude/env_default.h: "loadaddr="__stringify(CONFIG_LOADADDR)
"\0"Indeed - I missed this one. I only checked C and asm sources, and cound not find a single reference...
It's also contained in LARGE amount of config files of course :(
Quoted 9 lines
So
arch/blackfin/include/asm/config.handinclude/env_default.hare the only locateions that refer toCONFIG_LOADADDR? This should indeed be cleaned up. [But this is not related to your patch any longer.]Best regards,
Wolfgang Denk
Best regards,
Marek Vasut
PATCH 6/6 arm: mx5: Add support for DENX M53EVK
Dear Marek Vasut,
Quoted 9 lines · In message <[email protected]> you wrote:
Quoted 7 lines
Quoted 5 lines
Quoted 2 lines
What do we need
CONFIG_LOADADDRfor?For this stuff, the "loadaddr" env variable. I have kinda idea to kill this and unify it to
CONFIG_SYS_LOAD_ADDRall around, what do you think?Why CONFIG_SYS_* ?
What do you suggest then?
If we unify it, then we should use CONFIG_LOADADDR - this is a user defined preference setting, not really a system / hardware dependent one.
Quoted 8 lines
Quoted 6 lines
Quoted 3 lines
include/env_default.h:#ifdef
CONFIG_LOADADDRinclude/env_default.h: "loadaddr="__stringify(CONFIG_LOADADDR)
"\0"Indeed - I missed this one. I only checked C and asm sources, and cound not find a single reference...
It's also contained in LARGE amount of config files of course :(
Yes, they set it for this single place where it gets used.
Best regards,
Wolfgang Denk
Signature · 4 lines
DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de Successful and fortunate crime is called virtue. - Seneca
PATCH 6/6 arm: mx5: Add support for DENX M53EVK
Dear Wolfgang Denk,
Quoted 9 lines
Dear Marek Vasut,
Quoted 6 lines · In message <[email protected]> you wrote:
Quoted 6 lines
Quoted 5 lines
Quoted 1 line
What do we need
CONFIG_LOADADDRfor?For this stuff, the "loadaddr" env variable. I have kinda idea to kill this and unify it to
CONFIG_SYS_LOAD_ADDRall around, what do you think?
Quoted 7 lines
Quoted 3 lines
Quoted 1 line
Why CONFIG_SYS_* ?
What do you suggest then?
If we unify it, then we should use
CONFIG_LOADADDR- this is a user defined preference setting, not really a system / hardware dependent one.
OK
Quoted 10 lines
Quoted 8 lines
Quoted 6 lines
Quoted 3 lines
include/env_default.h:#ifdef
CONFIG_LOADADDRinclude/env_default.h: "loadaddr="__stringify(CONFIG_LOADADDR)
"\0"Indeed - I missed this one. I only checked C and asm sources, and cound not find a single reference...
It's also contained in LARGE amount of config files of course :(
Yes, they set it for this single place where it gets used.
Do you want me to fix it or will you ?
Best regards,
Marek Vasut
PATCH 6/6 arm: mx5: Add support for DENX M53EVK
Dear Marek,
Quoted 4 lines · In message <[email protected]> you wrote:
Quoted 2 lines
Yes, they set it for this single place where it gets used.
Do you want me to fix it or will you ?
If you have some time left... thanks in advance.
Best regards,
Wolfgang Denk
Signature · 6 lines
DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de Conceptual integrity in turn dictates that the design must proceed from one mind, or from a very small number of agreeing resonant minds. - Frederick Brooks Jr., "The Mythical Man Month"
PATCH 6/6 arm: mx5: Add support for DENX M53EVK
Dear Wolfgang Denk,
Quoted 8 lines
Dear Marek,
Quoted 3 lines · In message <[email protected]> you wrote:
Quoted 1 line
Yes, they set it for this single place where it gets used.
Do you want me to fix it or will you ?
If you have some time left... thanks in advance.
Looking through the code, it seems CONFIG_SYS_LOAD_ADDR is preferred over CONFIG_LOADADDR, so this should be fixed on all places then ? Just to be clear on this.
Best regards,
Marek Vasut
PATCH 6/6 arm: mx5: Add support for DENX M53EVK
Dear Marek,
Quoted 4 lines · In message <[email protected]> you wrote:
Looking through the code, it seems
CONFIG_SYS_LOAD_ADDRis preferred overCONFIG_LOADADDR, so this should be fixed on all places then ? Just to be clear on this.
CONFIG_SYS_LOAD_ADDR is _used_, but not preferred ;-)
Yes, I think we should eliminate CONFIG_SYS_LOAD_ADDR and globally replace it by CONFIG_LOADADDR.
Best regards,
Wolfgang Denk
Signature · 6 lines
DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
There is is no reason for any individual to have a computer in their
home. -- Ken Olsen (President of Digital Equipment Corporation),
Convention of the World Future Society, in Boston, 1977PATCH 6/6 arm: mx5: Add support for DENX M53EVK
Dear Wolfgang Denk,
Quoted 11 lines
Dear Marek,
Quoted 3 lines · In message <[email protected]> you wrote:
Looking through the code, it seems
CONFIG_SYS_LOAD_ADDRis preferred overCONFIG_LOADADDR, so this should be fixed on all places then ? Just to be clear on this.
CONFIG_SYS_LOAD_ADDRis _used_, but not preferred ;-)Yes, I think we should eliminate
CONFIG_SYS_LOAD_ADDRand globally replace it byCONFIG_LOADADDR.
EErgh ... for example tegra30-common.h uses both CONFIG_LOADADDR and CONFIG_SYS_LOAD_ADDR configured to different addresses.
Simon, we'd like to unify CONFIG_SYS_LOAD_ADDR and CONFIG_LOADADDR, why does tegra use different addresses for those two symbols? Can you fix it?
Best regards,
Marek Vasut
PATCH 6/6 arm: mx5: Add support for DENX M53EVK
In reply to Marek VasutDear Marek Vasut,
Quoted 517 lines · Marek Vasut
Add basic support for the DENX M53EVK board. Currently supported is the MMC, Ethernet, I2C.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>Cc Wolfgang Denk <[email protected]>--- MAINTAINERS | 1 + board/denx/m53evk/Makefile | 40 ++++ board/denx/m53evk/imximage.cfg | 108 +++++++++++ board/denx/m53evk/m53evk.c | 410 +++++++++++++++++++++++++++++++++++++++++ boards.cfg | 1 + include/configs/m53evk.h | 252 +++++++++++++++++++++++++ 6 files changed, 812 insertions(+) create mode 100644 board/denx/m53evk/Makefile create mode 100644 board/denx/m53evk/imximage.cfg create mode 100644 board/denx/m53evk/m53evk.c create mode 100644 include/configs/m53evk.hMAINTAINERS +1
-0@@ -954,6 +954,7 @@ Marek Vasut <[email protected]> mx23_olinuxino i.MX23 m28evk i.MX28 sc_sps_1 i.MX28 + m53evk i.MX53 Hugo Villeneuve <[email protected]>board/denx/m53evk/Makefile new file +40
-0@@ -0,0 +1,40 @@ +# +# DENX M53EVK +# Copyright (C) 2012-2013 Marek Vasut <[email protected]> +# +# 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 $(TOPDIR)/config.mk + +LIB = $(obj)lib$(BOARD).o + +COBJS := m53evk.o + +SRCS := $(COBJS:.o=.c) +OBJS := $(addprefix $(obj),$(COBJS)) + +$(LIB): $(obj).depend $(OBJS) + $(call cmd_link_o_target, $(OBJS)) + +######################################################################### + +# defines $(obj).depend target +include $(SRCTREE)/rules.mk + +sinclude $(obj).depend + +#########################################################################board/denx/m53evk/imximage.cfg new file +108
-0@@ -0,0 +1,108 @@ +/* + * DENX M53 DRAM init values + * Copyright (C) 2012-2013 Marek Vasut <[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.imxmage for more details about how-to configure + * and create imximage boot image + * + * The syntax is taken as close as possible with the kwbimage + */ +/* image version */ +IMAGE_VERSION 2 + +/* + * Boot Device : one of + * spi, sd, nand + */ +BOOT_FROM nand + +/* + * Device Configuration Data (DCD) + * + * Each entry must have the format: + * Addr-type Address Value + * + * where: + * Addr-type register length (1,2 or 4 bytes) + * Address absolute address of the register + * value value to be stored in the register + */ +DATA 4 0x53fa86f4 0x00000000 /* GRP_DDRMODE_CTL */ +DATA 4 0x53fa8714 0x00000000 /* GRP_DDRMODE */ +DATA 4 0x53fa86fc 0x00000000 /* GRP_DDRPKE */ +DATA 4 0x53fa8724 0x04000000 /* GRP_DDR_TYPE */ + +DATA 4 0x53fa872c 0x00300000 /* GRP_B3DS */ +DATA 4 0x53fa8554 0x00300000 /* DRAM_DQM3 */ +DATA 4 0x53fa8558 0x00300040 /* DRAM_SDQS3 */ + +DATA 4 0x53fa8728 0x00300000 /* GRP_B2DS */ +DATA 4 0x53fa8560 0x00300000 /* DRAM_DQM2 */ +DATA 4 0x53fa8568 0x00300040 /* DRAM_SDQS2 */ + +DATA 4 0x53fa871c 0x00300000 /* GRP_B1DS */ +DATA 4 0x53fa8594 0x00300000 /* DRAM_DQM1 */ +DATA 4 0x53fa8590 0x00300040 /* DRAM_SDQS1 */ + +DATA 4 0x53fa8718 0x00300000 /* GRP_B0DS */ +DATA 4 0x53fa8584 0x00300000 /* DRAM_DQM0 */ +DATA 4 0x53fa857c 0x00300040 /* DRAM_SDQS0 */ + +DATA 4 0x53fa8578 0x00300000 /* DRAM_SDCLK_0 */ +DATA 4 0x53fa8570 0x00300000 /* DRAM_SDCLK_1 */ + +DATA 4 0x53fa8574 0x00300000 /* DRAM_CAS */ +DATA 4 0x53fa8588 0x00300000 /* DRAM_RAS */ +DATA 4 0x53fa86f0 0x00300000 /* GRP_ADDDS */ +DATA 4 0x53fa8720 0x00300000 /* GRP_CTLDS */ + +DATA 4 0x53fa8564 0x00300040 /* DRAM_SDODT1 */ +DATA 4 0x53fa8580 0x00300040 /* DRAM_SDODT0 */ + +/* ESDCTL */ +DATA 4 0x63fd9088 0x32383535 +DATA 4 0x63fd9090 0x40383538 +DATA 4 0x63fd907c 0x0136014d +DATA 4 0x63fd9080 0x01510141 + +DATA 4 0x63fd9018 0x00011740 +DATA 4 0x63fd9000 0xc3190000 +DATA 4 0x63fd900c 0x555952e3 +DATA 4 0x63fd9010 0xb68e8b63 +DATA 4 0x63fd9014 0x01ff00db +DATA 4 0x63fd902c 0x000026d2 +DATA 4 0x63fd9030 0x009f0e21 +DATA 4 0x63fd9008 0x12273030 +DATA 4 0x63fd9004 0x0002002d +DATA 4 0x63fd901c 0x00008032 +DATA 4 0x63fd901c 0x00008033 +DATA 4 0x63fd901c 0x00028031 +DATA 4 0x63fd901c 0x092080b0 +DATA 4 0x63fd901c 0x04008040 +DATA 4 0x63fd901c 0x0000803a +DATA 4 0x63fd901c 0x0000803b +DATA 4 0x63fd901c 0x00028039 +DATA 4 0x63fd901c 0x09208138 +DATA 4 0x63fd901c 0x04008048 +DATA 4 0x63fd9020 0x00001800 +DATA 4 0x63fd9040 0x04b80003 +DATA 4 0x63fd9058 0x00022227 +DATA 4 0x63fd901c 0x00000000board/denx/m53evk/m53evk.c new file +314
-0@@ -0,0 +1,410 @@ +/* +* DENX M53 module +* +* Copyright (C) 2012-2013 Marek Vasut <[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> +#include <asm/io.h> +#include <asm/arch/imx-regs.h> +#include <asm/arch/mx5x_pins.h> +#include <asm/arch/sys_proto.h> +#include <asm/arch/crm_regs.h> +#include <asm/arch/clock.h> +#include <asm/arch/iomux.h> +#include <asm/arch/spl.h> +#include <asm/errno.h> +#include <netdev.h> +#include <i2c.h> +#include <mmc.h> +#include <spl.h> +#include <fsl_esdhc.h> +#include <asm/gpio.h> +#include <usb/ehci-fsl.h> + +DECLARE_GLOBAL_DATA_PTR ; + +int dram_init (void ) +{ +u32 size1, size2; + + size1 =get_ram_size ((void *)PHYS_SDRAM_1 ,PHYS_SDRAM_1_SIZE ); + size2 =get_ram_size ((void *)PHYS_SDRAM_2 ,PHYS_SDRAM_2_SIZE ); + + gd->ram_size = size1 + size2; + +return 0 ; +} +void dram_init_banksize (void ) +{ + gd->bd->bi_dram[0 ].start =PHYS_SDRAM_1 ; + gd->bd->bi_dram[0 ].size =PHYS_SDRAM_1_SIZE ; + + gd->bd->bi_dram[1 ].start =PHYS_SDRAM_2 ; + gd->bd->bi_dram[1 ].size =PHYS_SDRAM_2_SIZE ; +} + +u32 get_board_rev (void ) +{ +struct iim_regs *iim = (struct iim_regs *)IMX_IIM_BASE ; +struct fuse_bank *bank = &iim->bank[0 ]; +struct fuse_bank0_regs *fuse = + (struct fuse_bank0_regs *)bank->fuse_regs; +int rev =readl (&fuse->gp[6 ]); + +return (get_cpu_rev () & ~(0xF <<8 )) | (rev &0xF ) <<8 ; +} + +static void setup_iomux_uart (void ) +{ +mxc_request_iomux (MX53_PIN_ATA_BUFFER_EN ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DMARQ ,IOMUX_CONFIG_ALT3 ); + +mxc_iomux_set_pad (MX53_PIN_ATA_BUFFER_EN , +PAD_CTL_SRE_FAST |PAD_CTL_DRV_HIGH | +PAD_CTL_PUE_PULL |PAD_CTL_PKE_ENABLE | +PAD_CTL_HYS_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DMARQ , +PAD_CTL_SRE_FAST |PAD_CTL_DRV_HIGH | +PAD_CTL_PUE_PULL |PAD_CTL_PKE_ENABLE | +PAD_CTL_HYS_ENABLE ); + +mxc_iomux_set_input (MX53_UART2_IPP_UART_RXD_MUX_SELECT_INPUT ,0x3 ); +} + +#ifdef CONFIG_USB_EHCI_MX5 +int board_ehci_hcd_init (int port) +{ +if (port ==0 ) { +/* USB OTG PWRON */ +mxc_request_iomux (MX53_PIN_GPIO_4 ,IOMUX_CONFIG_ALT1 ); +mxc_iomux_set_pad (MX53_PIN_GPIO_4 , +PAD_CTL_PKE_ENABLE | +PAD_CTL_100K_PD | +PAD_CTL_DRV_HIGH + ); +gpio_direction_output (IOMUX_TO_GPIO (MX53_PIN_GPIO_4 ),0 ); + +/* USB OTG Over Current */ +mxc_request_iomux (MX53_PIN_GPIO_18 ,IOMUX_CONFIG_ALT1 ); +mxc_iomux_set_input (MX53_USBOH3_IPP_IND_OTG_OC_SELECT_INPUT ,1 ); + }else if (port ==1 ) { +/* USB Host PWRON */ +mxc_request_iomux (MX53_PIN_GPIO_2 ,IOMUX_CONFIG_ALT1 ); +mxc_iomux_set_pad (MX53_PIN_GPIO_2 , +PAD_CTL_PKE_ENABLE | +PAD_CTL_100K_PD | +PAD_CTL_DRV_HIGH + ); +gpio_direction_output (IOMUX_TO_GPIO (MX53_PIN_GPIO_2 ),0 ); + +/* USB Host Over Current */ +mxc_request_iomux (MX53_PIN_GPIO_3 ,IOMUX_CONFIG_ALT6 ); +mxc_iomux_set_input (MX53_USBOH3_IPP_IND_UH1_OC_SELECT_INPUT ,1 ); + } + +return 0 ; +} +#endif + +static void setup_iomux_fec (void ) +{ +/* MDIO IOMUX */ +mxc_request_iomux (MX53_PIN_FEC_MDIO ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_FEC_MDC ,IOMUX_CONFIG_ALT0 ); + +/* FEC 0 IOMUX */ +mxc_request_iomux (MX53_PIN_FEC_CRS_DV ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_FEC_REF_CLK ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_FEC_RX_ER ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_FEC_TX_EN ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_FEC_RXD0 ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_FEC_RXD1 ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_FEC_TXD0 ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_FEC_TXD1 ,IOMUX_CONFIG_ALT0 ); + +/* FEC 1 IOMUX */ +mxc_request_iomux (MX53_PIN_KEY_COL0 ,IOMUX_CONFIG_ALT6 );/* RXD3 */ +mxc_request_iomux (MX53_PIN_KEY_ROW0 ,IOMUX_CONFIG_ALT6 );/* TX_ER */ +mxc_request_iomux (MX53_PIN_KEY_COL1 ,IOMUX_CONFIG_ALT6 );/* RX_CLK */ +mxc_request_iomux (MX53_PIN_KEY_ROW1 ,IOMUX_CONFIG_ALT6 );/* COL */ +mxc_request_iomux (MX53_PIN_KEY_COL2 ,IOMUX_CONFIG_ALT6 );/* RXD2 */ +mxc_request_iomux (MX53_PIN_KEY_ROW2 ,IOMUX_CONFIG_ALT6 );/* TXD2 */ +mxc_request_iomux (MX53_PIN_KEY_COL3 ,IOMUX_CONFIG_ALT6 );/* CRS */ +mxc_request_iomux (MX53_PIN_GPIO_19 ,IOMUX_CONFIG_ALT6 );/* TXD3 */ + +/* MDIO PADs */ +mxc_iomux_set_pad (MX53_PIN_FEC_MDIO , +PAD_CTL_HYS_ENABLE |PAD_CTL_DRV_HIGH | +PAD_CTL_PUE_PULL |PAD_CTL_PKE_ENABLE | +PAD_CTL_22K_PU |PAD_CTL_ODE_OPENDRAIN_ENABLE ); +mxc_iomux_set_input (MX53_FEC_FEC_MDI_SELECT_INPUT ,0x1 ); +mxc_iomux_set_pad (MX53_PIN_FEC_MDC ,PAD_CTL_DRV_HIGH ); + +/* FEC 0 PADs */ +mxc_iomux_set_pad (MX53_PIN_FEC_CRS_DV , +PAD_CTL_HYS_ENABLE |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_FEC_REF_CLK , +PAD_CTL_HYS_ENABLE |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_FEC_RX_ER , +PAD_CTL_HYS_ENABLE |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_FEC_TX_EN ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_FEC_RXD0 , +PAD_CTL_HYS_ENABLE |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_FEC_RXD1 , +PAD_CTL_HYS_ENABLE |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_FEC_TXD0 ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_FEC_TXD1 ,PAD_CTL_DRV_HIGH ); + +/* FEC 1 PADs */ +mxc_iomux_set_pad (MX53_PIN_KEY_COL0 ,/* RXD3 */ +PAD_CTL_HYS_ENABLE |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_KEY_ROW0 ,/* TX_ER */ +PAD_CTL_HYS_ENABLE |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_KEY_COL1 ,/* RX_CLK */ +PAD_CTL_HYS_ENABLE |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_KEY_ROW1 ,/* COL */ +PAD_CTL_HYS_ENABLE |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_KEY_COL2 ,/* RXD2 */ +PAD_CTL_HYS_ENABLE |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_KEY_ROW2 ,/* TXD2 */ +PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_KEY_COL3 ,/* CRS */ +PAD_CTL_HYS_ENABLE |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_GPIO_19 ,/* TXD3 */ +PAD_CTL_DRV_HIGH ); +} + +#ifdef CONFIG_FSL_ESDHC +struct fsl_esdhc_cfg esdhc_cfg = { +MMC_SDHC1_BASE_ADDR , +}; + +int board_mmc_getcd (struct mmc *mmc) +{ +mxc_request_iomux (MX53_PIN_GPIO_1 ,IOMUX_CONFIG_ALT1 ); +gpio_direction_input (IMX_GPIO_NR (1 ,1 )); + +return !gpio_get_value (IMX_GPIO_NR (1 ,1 )); +} + +int board_mmc_init (bd_t *bis) +{ + esdhc_cfg.sdhc_clk =mxc_get_clock (MXC_ESDHC_CLK ); + +mxc_request_iomux (MX53_PIN_SD1_CMD ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_SD1_CLK ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_SD1_DATA0 , +IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_SD1_DATA1 , +IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_SD1_DATA2 , +IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_SD1_DATA3 , +IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_EIM_DA13 , +IOMUX_CONFIG_ALT1 ); + +mxc_iomux_set_pad (MX53_PIN_SD1_CMD , +PAD_CTL_HYS_ENABLE |PAD_CTL_DRV_HIGH | +PAD_CTL_PUE_PULL |PAD_CTL_PKE_ENABLE | +PAD_CTL_HYS_ENABLE |PAD_CTL_100K_PU ); +mxc_iomux_set_pad (MX53_PIN_SD1_CLK , +PAD_CTL_PUE_PULL |PAD_CTL_PKE_ENABLE | +PAD_CTL_HYS_ENABLE |PAD_CTL_47K_PU | +PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_SD1_DATA0 , +PAD_CTL_HYS_ENABLE |PAD_CTL_DRV_HIGH | +PAD_CTL_PUE_PULL |PAD_CTL_PKE_ENABLE | +PAD_CTL_HYS_ENABLE |PAD_CTL_47K_PU ); +mxc_iomux_set_pad (MX53_PIN_SD1_DATA1 , +PAD_CTL_HYS_ENABLE |PAD_CTL_DRV_HIGH | +PAD_CTL_PUE_PULL |PAD_CTL_PKE_ENABLE | +PAD_CTL_HYS_ENABLE |PAD_CTL_47K_PU ); +mxc_iomux_set_pad (MX53_PIN_SD1_DATA2 , +PAD_CTL_HYS_ENABLE |PAD_CTL_DRV_HIGH | +PAD_CTL_PUE_PULL |PAD_CTL_PKE_ENABLE | +PAD_CTL_HYS_ENABLE |PAD_CTL_47K_PU ); +mxc_iomux_set_pad (MX53_PIN_SD1_DATA3 , +PAD_CTL_HYS_ENABLE |PAD_CTL_DRV_HIGH | +PAD_CTL_PUE_PULL |PAD_CTL_PKE_ENABLE | +PAD_CTL_HYS_ENABLE |PAD_CTL_47K_PU ); + +/* GPIO 2_31 is SD power */ +mxc_request_iomux (MX53_PIN_EIM_EB3 ,IOMUX_CONFIG_ALT1 ); +gpio_direction_output (IMX_GPIO_NR (2 ,31 ),0 ); + +return fsl_esdhc_initialize (bis, &esdhc_cfg); +} +#endif + +static void setup_iomux_i2c (void ) +{ +mxc_request_iomux (MX53_PIN_EIM_D16 , +IOMUX_CONFIG_ALT5 |IOMUX_CONFIG_SION ); +mxc_request_iomux (MX53_PIN_EIM_EB2 , +IOMUX_CONFIG_ALT5 |IOMUX_CONFIG_SION ); + +mxc_iomux_set_pad (MX53_PIN_EIM_D16 , +PAD_CTL_SRE_FAST |PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE | +PAD_CTL_PUE_PULL | +PAD_CTL_ODE_OPENDRAIN_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_EIM_EB2 , +PAD_CTL_SRE_FAST |PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE | +PAD_CTL_PUE_PULL | +PAD_CTL_ODE_OPENDRAIN_ENABLE ); + +mxc_iomux_set_input (MX53_I2C2_IPP_SDA_IN_SELECT_INPUT ,0x1 ); +mxc_iomux_set_input (MX53_I2C2_IPP_SCL_IN_SELECT_INPUT ,0x1 ); +} + +static void setup_iomux_nand (void ) +{ +mxc_request_iomux (MX53_PIN_NANDF_WE_B ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_RE_B ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_CLE ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_ALE ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_WP_B ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_RB0 ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_CS0 ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_ATA_DATA0 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA1 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA2 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA3 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA4 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA5 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA6 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA7 ,IOMUX_CONFIG_ALT3 ); + +mxc_iomux_set_pad (MX53_PIN_NANDF_WE_B ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_NANDF_RE_B ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_NANDF_CLE ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_NANDF_ALE ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_NANDF_WP_B ,PAD_CTL_PUE_PULL | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_NANDF_RB0 ,PAD_CTL_PUE_PULL | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_NANDF_CS0 ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA0 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA1 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA2 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); + mxc_iomux_set_pad(MX53_PIN_ATA_DATA3, PAD_CTL_DRV_HIGH |
board/denx/m53evk/m53evk.c line 324
Quoted 10 lines
board/denx/m53evk/m53evk.c new file +10
-0+ PAD_CTL_100K_PU | PAD_CTL_PKE_ENABLE); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA4 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA5 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA6 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA7 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +}
Like for 5/6, why is this function needed? When booting from NAND, the boot ROM sets the NAND pads appropriately itself (see 7.5.2.4), so unless it reverts those changes to the IOMUX reset values afterwards, there is no need to set the NAND IOMUX again. It's not as if the board were using NAND after having been booted from another source (like mx53ard does).
Or perhaps you are planning to add more boot sources later, in which case there are configs missing for M4IF and WEIM (see mx53ard's setup_iomux_nand()).
Quoted 369 lines
board/denx/m53evk/m53evk.c new file +86
-0+ +static void m53_set_clock (void ) +{ +int ret; +const uint32_t ref_clk =MXC_HCLK ; +const uint32_t dramclk =400 ; +uint32_t cpuclk; + +mxc_request_iomux (MX53_PIN_GPIO_10 ,IOMUX_CONFIG_GPIO ); +mxc_iomux_set_pad (MX53_PIN_GPIO_10 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +gpio_direction_input (IOMUX_TO_GPIO (MX53_PIN_GPIO_10 )); + +/* GPIO10 selects modules' CPU speed, 1 = 1200MHz ; 0 = 800MHz */ + cpuclk =gpio_get_value (IOMUX_TO_GPIO (MX53_PIN_GPIO_10 )) ?1200 :800 ; + + ret =mxc_set_clock (ref_clk, cpuclk,MXC_ARM_CLK ); +if (ret) +printf ("CPU: Switch CPU clock to %dMHz failed\n" , cpuclk); + + ret =mxc_set_clock (ref_clk, dramclk,MXC_PERIPH_CLK ); +if (ret) { +printf ("CPU: Switch peripheral clock to %dMHz failed\n" , + dramclk); + } + + ret =mxc_set_clock (ref_clk, dramclk,MXC_DDR_CLK ); +if (ret) +printf ("CPU: Switch DDR clock to %dMHz failed\n" , dramclk); +} + +int board_early_init_f (void ) +{ +setup_iomux_uart (); +setup_iomux_fec (); +setup_iomux_i2c (); +setup_iomux_nand (); + +m53_set_clock (); + +return 0 ; +} + +/* +* Do not overwrite the console +* Use always serial for U-Boot console +*/ +int overwrite_console (void ) +{ +return 1 ; +} + +int board_init (void ) +{ + gd->bd->bi_boot_params =PHYS_SDRAM_1 +0x100 ; + +mxc_set_sata_internal_clock (); + +mxc_set_clock (0 ,33 ,MXC_NFC_CLK ); +enable_nfc_clk (1 ); + +return 0 ; +} + +int checkboard (void ) +{ +puts ("Board: DENX M53EVK\n" ); + +return 0 ; +} + +/* +* NAND SPL +*/ +#ifdef CONFIG_SPL_BUILD +void spl_board_init (void ) +{ +m53_set_clock (); +setup_iomux_nand (); +} + +u32 spl_boot_device (void ) +{ +return BOOT_DEVICE_NAND ; +} +#endif boards.cfg +0
-0@@ -246,6 +246,7 @@ am335x_evm_usbspl arm armv7am335x ti
ti814x_evmarm armv7 ti814x ti
am33xx
pcm051 arm armv7 pcm051
phytec am33xx pcm051
highbank arm armv7 highbank -
highbank
+m53evk arm armv7 m53evk
denx mx5 m53evk:IMX_CONFIG=board/denx/m53evk/imximage.cfg
mx51_efikamxarm armv7mx51_efikamx
genesi mx5boards.cfg +0
-0mx51_efikamx:MACH_TYPE=MACH_TYPE_MX51_EFIKAMX,IMX_CONFIG=board/genesi/mx51_efikamx/imximage_mx.cfg mx51_efikasb arm armv7 mx51_efikamx genesi mx5 mx51_efikamx:MACH_TYPE=MACH_TYPE_MX51_EFIKASB,IMX_CONFIG=board/genesi/mx51_efikamx/imximage_sb.cfg mx51evk arm armv7 mx51evk freescale mx5 mx51evk:IMX_CONFIG=board/freescale/mx51evk/imximage.cfginclude/configs/m53evk.h new file +234
-0@@ -0,0 +1,252 @@ +/* +* DENX M53 configuration +* Copyright (C) 2012-2013 Marek Vasut <[email protected]> +* +* 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 +*/ + +#ifndef __M53EVK_CONFIG_H__ +#define __M53EVK_CONFIG_H__ + +#define CONFIG_MX53 +#define CONFIG_MXC_GPIO +#define CONFIG_SYS_HZ 1000 + +#include <asm/arch/imx-regs.h> + +#define CONFIG_DISPLAY_CPUINFO +#define CONFIG_BOARD_EARLY_INIT_F +#define CONFIG_REVISION_TAG +#define CONFIG_SYS_NO_FLASH + +/* +* U-Boot Commands +*/ +#include <config_cmd_default.h> +#define CONFIG_DISPLAY_BOARDINFO +#define CONFIG_DOS_PARTITION + +#define CONFIG_CMD_DHCP +#define CONFIG_CMD_EXT2 +#define CONFIG_CMD_FAT +#define CONFIG_CMD_FAT +#define CONFIG_CMD_I2C +#define CONFIG_CMD_MII +#define CONFIG_CMD_MMC +#define CONFIG_CMD_NAND +#define CONFIG_CMD_NET +#define CONFIG_CMD_PING +#define CONFIG_CMD_SATA +#define CONFIG_CMD_USB + +/* +* Memory configurations +*/ +#define CONFIG_NR_DRAM_BANKS 2 +#define PHYS_SDRAM_1 CSD0_BASE_ADDR +#define PHYS_SDRAM_1_SIZE (512 * 1024 * 1024) +#define PHYS_SDRAM_2 CSD1_BASE_ADDR +#define PHYS_SDRAM_2_SIZE (512 * 1024 * 1024) +#define PHYS_SDRAM_SIZE (PHYS_SDRAM_1_SIZE + PHYS_SDRAM_2_SIZE) +#define CONFIG_SYS_MALLOC_LEN (10 * 1024 * 1024) +#define CONFIG_SYS_MEMTEST_START 0x70010000 +#define CONFIG_SYS_MEMTEST_END 0x70020000 + +#define CONFIG_SYS_SDRAM_BASE (PHYS_SDRAM_1) +#define CONFIG_SYS_INIT_RAM_ADDR (IRAM_BASE_ADDR) +#define CONFIG_SYS_INIT_RAM_SIZE (IRAM_SIZE) + +#define CONFIG_SYS_INIT_SP_OFFSET \ + (CONFIG_SYS_INIT_RAM_SIZE - GENERATED_GBL_DATA_SIZE) +#define CONFIG_SYS_INIT_SP_ADDR \ + (CONFIG_SYS_INIT_RAM_ADDR + CONFIG_SYS_INIT_SP_OFFSET) + +#define CONFIG_SYS_TEXT_BASE 0x71000000 + +/* +* U-Boot general configurations +*/ +#define CONFIG_SYS_LONGHELP +#define CONFIG_SYS_PROMPT "=> " +#define CONFIG_SYS_CBSIZE 1024/* Console I/O buffer size */ +#define CONFIG_SYS_PBSIZE \ + (CONFIG_SYS_CBSIZE + sizeof(CONFIG_SYS_PROMPT) + 16) +/* Print buffer size */ +#define CONFIG_SYS_MAXARGS 32/* Max number of command args */ +#define CONFIG_SYS_BARGSIZE CONFIG_SYS_CBSIZE +/* Boot argument buffer size */ +#define CONFIG_VERSION_VARIABLE /* U-BOOT version */ +#define CONFIG_AUTO_COMPLETE /* Command auto complete */ +#define CONFIG_CMDLINE_EDITING /* Command history etc */ +#define CONFIG_SYS_HUSH_PARSER + +/* +* Serial Driver +*/ +#define CONFIG_MXC_UART +#define CONFIG_MXC_UART_BASE UART2_BASE +#define CONFIG_CONS_INDEX 1 +#define CONFIG_BAUDRATE 115200 + +/* +* MMC Driver +*/ +#ifdef CONFIG_CMD_MMC +#define CONFIG_MMC +#define CONFIG_GENERIC_MMC +#define CONFIG_FSL_ESDHC +#define CONFIG_SYS_FSL_ESDHC_ADDR 0 +#define CONFIG_SYS_FSL_ESDHC_NUM 1 +#endif + +/* +* NAND +*/ +#define CONFIG_ENV_SIZE (16 * 1024) +#ifdef CONFIG_CMD_NAND +#define CONFIG_SYS_MAX_NAND_DEVICE 1 +#define CONFIG_SYS_NAND_BASE NFC_BASE_ADDR_AXI +#define CONFIG_NAND_MXC +#define CONFIG_MXC_NAND_REGS_BASE NFC_BASE_ADDR_AXI +#define CONFIG_MXC_NAND_IP_REGS_BASE NFC_BASE_ADDR +#define CONFIG_SYS_NAND_LARGEPAGE +#define CONFIG_MXC_NAND_HWECC +#define CONFIG_SYS_NAND_USE_FLASH_BBT + +/* Environment is in NAND */ +#define CONFIG_ENV_IS_IN_NAND +#define CONFIG_ENV_SIZE_REDUND CONFIG_ENV_SIZE +#define CONFIG_ENV_SECT_SIZE (128 * 1024) +#define CONFIG_ENV_RANGE (512 * 1024) +#define CONFIG_ENV_OFFSET 0x100000 +#define CONFIG_ENV_OFFSET_REDUND \ + (CONFIG_ENV_OFFSET + CONFIG_ENV_RANGE) + +#define CONFIG_CMD_UBI +#define CONFIG_CMD_UBIFS +#define CONFIG_CMD_MTDPARTS +#define CONFIG_RBTREE +#define CONFIG_LZO +#define CONFIG_MTD_DEVICE +#define CONFIG_MTD_PARTITIONS +#define MTDIDS_DEFAULT "nand0=mxc-nand" +#define MTDPARTS_DEFAULT \ + "mtdparts=mxc-nand:" \ + "1m(bootloader)ro," \ + "512k(environment)," \ + "512k(redundant-environment)," \ + "4m(kernel)," \ + "128k(fdt)," \ + "8m(ramdisk)," \ + "-(filesystem)" +#else +#define CONFIG_ENV_IS_NOWHERE +#endif + +/* +* Ethernet on SOC (FEC) +*/ +#ifdef CONFIG_CMD_NET +#define CONFIG_FEC_MXC +#define CONFIG_ETHPRIME "FEC0" +#define IMX_FEC_BASE FEC_BASE_ADDR +#define CONFIG_FEC_MXC_PHYADDR 0x0 +#define CONFIG_MII +#define CONFIG_DISCOVER_PHY +#define CONFIG_FEC_XCV_TYPE RMII +#define CONFIG_PHYLIB +#define CONFIG_PHY_MICREL +#define CONFIG_ARP_TIMEOUT 200UL +#endif + +/* +* I2C +*/ +#ifdef CONFIG_CMD_I2C +#define CONFIG_HARD_I2C +#define CONFIG_I2C_MXC +#define CONFIG_SYS_I2C_BASE I2C2_BASE_ADDR +#define CONFIG_SYS_I2C_SPEED 100000 +#endif + +/* +* USB +*/ +#ifdef CONFIG_CMD_USB +#define CONFIG_USB_EHCI +#define CONFIG_USB_EHCI_MX5 +#define CONFIG_USB_STORAGE +#define CONFIG_USB_HOST_ETHER +#define CONFIG_USB_ETHER_ASIX +#define CONFIG_USB_ETHER_SMSC95XX +#define CONFIG_MXC_USB_PORT 1 +#define CONFIG_MXC_USB_PORTSC (PORT_PTS_UTMI | PORT_PTS_PTW) +#define CONFIG_MXC_USB_FLAGS 0 +#endif + +/* +* SATA +*/ +#ifdef CONFIG_CMD_SATA +#define CONFIG_DWC_AHSATA +#define CONFIG_SYS_SATA_MAX_DEVICE 1 +#define CONFIG_DWC_AHSATA_PORT_ID 0 +#define CONFIG_DWC_AHSATA_BASE_ADDR SATA_BASE_ADDR +#define CONFIG_LBA48 +#define CONFIG_LIBATA +#endif + +/* +* Boot Linux +*/ +#define CONFIG_CMDLINE_TAG +#define CONFIG_INITRD_TAG +#define CONFIG_SETUP_MEMORY_TAGS +#define CONFIG_BOOTDELAY 3 +#define CONFIG_BOOTFILE "uImage" +#define CONFIG_BOOTARGS "console=ttymxc1,115200" +#define CONFIG_BOOTCOMMAND "run bootcmd_net" +#define CONFIG_LOADADDR 0x70800000 +#define CONFIG_SYS_LOAD_ADDR CONFIG_LOADADDR +#define CONFIG_OF_LIBFDT + +/* +* NAND SPL +*/ +#define CONFIG_SPL +#define CONFIG_SPL_FRAMEWORK +#define CONFIG_SPL_BOARD_INIT +/* +* The 0x1000 offset must be present, otherwise board won't boot. +* The value (IRAM_BASE_ADDR) must be spelled out, otherwise linker won't link.
+ */include/configs/m53evk.h new file +17
-1+#define CONFIG_SPL_TEXT_BASE 0x70008000 +#define CONFIG_SPL_PAD_TO 0x8000 +#define CONFIG_SPL_STACK 0x70004000 +#define CONFIG_SPL_LIBCOMMON_SUPPORT +#define CONFIG_SPL_LIBGENERIC_SUPPORT +#define CONFIG_SPL_SERIAL_SUPPORT +#define CONFIG_SPL_GPIO_SUPPORT + +#define CONFIG_SPL_NAND_SUPPORT +#define CONFIG_SYS_NAND_U_BOOT_OFFS CONFIG_SPL_PAD_TO +#define CONFIG_SYS_NAND_PAGE_SIZE 2048 +#define CONFIG_SYS_NAND_OOBSIZE 64 +#define CONFIG_SYS_NAND_PAGE_COUNT 64 +#define CONFIG_SYS_NAND_SIZE (256 * 1024 * 1024) +#define CONFIG_SYS_NAND_BAD_BLOCK_POS 0 + +#endif /* __M53EVK_CONFIG_H__ */ --
Quoted 1 line
1.7.11.7
Best regards,
Beno?t
PATCH 6/6 arm: mx5: Add support for DENX M53EVK
Dear Beno?t Th?baudeau,
[...]
Quoted 38 lines
Quoted 38 lines
board/denx/m53evk/m53evk.c new file +38
-0+static void setup_iomux_nand (void ) +{ +mxc_request_iomux (MX53_PIN_NANDF_WE_B ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_RE_B ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_CLE ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_ALE ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_WP_B ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_RB0 ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_CS0 ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_ATA_DATA0 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA1 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA2 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA3 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA4 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA5 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA6 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA7 ,IOMUX_CONFIG_ALT3 ); + +mxc_iomux_set_pad (MX53_PIN_NANDF_WE_B ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_NANDF_RE_B ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_NANDF_CLE ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_NANDF_ALE ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_NANDF_WP_B ,PAD_CTL_PUE_PULL | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_NANDF_RB0 ,PAD_CTL_PUE_PULL | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_NANDF_CS0 ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA0 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA1 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA2 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA3 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA4 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); + mxc_iomux_set_pad(MX53_PIN_ATA_DATA5, PAD_CTL_DRV_HIGH |
Quoted 8 lines
board/denx/m53evk/m53evk.c line 324
Quoted 6 lines
board/denx/m53evk/m53evk.c new file +6
-0+ PAD_CTL_100K_PU | PAD_CTL_PKE_ENABLE); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA6 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA7 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +}Like for 5/6, why is this function needed? When booting from NAND
Where did you get the information the board can only boot from NAND? Did I mistakenly present it somewhere like that please?
Anyway, no, it can also boot from SD and USB at least.
Quoted 5 lines
the boot
ROM sets the NAND pads appropriately itself (see 7.5.2.4), so unless it reverts those changes to the IOMUX reset values afterwards, there is no need to set the NAND IOMUX again. It's not as if the board were using NAND after having been booted from another source (like mx53ard does).
I want to use NAND when booted from other sources, yes.
Quoted 3 lines
Or perhaps you are planning to add more boot sources later, in which case there are configs missing for M4IF and WEIM (see mx53ard's
setup_iomux_nand()).
Yes, but even without this configuration, the NAND works perfectly well when booted from SD. Am I missing something here?
[...]
PATCH 6/6 arm: mx5: Add support for DENX M53EVK
Dear Marek Vasut,
Quoted 51 lines · Marek Vasut
Dear Beno?t Th?baudeau,
[...]
Quoted 38 lines
Quoted 38 lines
board/denx/m53evk/m53evk.c new file +38
-0+static void setup_iomux_nand (void ) +{ +mxc_request_iomux (MX53_PIN_NANDF_WE_B ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_RE_B ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_CLE ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_ALE ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_WP_B ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_RB0 ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_NANDF_CS0 ,IOMUX_CONFIG_ALT0 ); +mxc_request_iomux (MX53_PIN_ATA_DATA0 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA1 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA2 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA3 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA4 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA5 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA6 ,IOMUX_CONFIG_ALT3 ); +mxc_request_iomux (MX53_PIN_ATA_DATA7 ,IOMUX_CONFIG_ALT3 ); + +mxc_iomux_set_pad (MX53_PIN_NANDF_WE_B ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_NANDF_RE_B ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_NANDF_CLE ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_NANDF_ALE ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_NANDF_WP_B ,PAD_CTL_PUE_PULL | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_NANDF_RB0 ,PAD_CTL_PUE_PULL | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_NANDF_CS0 ,PAD_CTL_DRV_HIGH ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA0 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA1 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA2 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA3 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA4 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); + mxc_iomux_set_pad(MX53_PIN_ATA_DATA5, PAD_CTL_DRV_HIGH |Quoted 8 lines
board/denx/m53evk/m53evk.c line 324
Quoted 6 lines
board/denx/m53evk/m53evk.c new file +6
-0+ PAD_CTL_100K_PU | PAD_CTL_PKE_ENABLE); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA6 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +mxc_iomux_set_pad (MX53_PIN_ATA_DATA7 ,PAD_CTL_DRV_HIGH | +PAD_CTL_100K_PU |PAD_CTL_PKE_ENABLE ); +}Like for 5/6, why is this function needed? When booting from NAND
Quoted 2 lines
Where did you get the information the board can only boot from NAND? Did I mistakenly present it somewhere like that please?
No, it's just suggested by the series, e.g. because imximage.cfg is NAND-specific.
Quoted 1 line
Anyway, no, it can also boot from SD and USB at least.
Then it's fine.
Quoted 7 lines
Quoted 5 lines
the boot
ROM sets the NAND pads appropriately itself (see 7.5.2.4), so unless it reverts those changes to the IOMUX reset values afterwards, there is no need to set the NAND IOMUX again. It's not as if the board were using NAND after having been booted from another source (like mx53ard does).I want to use NAND when booted from other sources, yes.
OK.
Quoted 6 lines
Quoted 3 lines
Or perhaps you are planning to add more boot sources later, in which case there are configs missing for M4IF and WEIM (see mx53ard's
setup_iomux_nand()).Yes, but even without this configuration, the NAND works perfectly well when booted from SD. Am I missing something here?
It's just a matter of precaution. If you think that you can't rely on the auto IOMUX or clock setup by the boot ROM, the same applies to M4IF and WEIM.
Best regards,
Beno?t
PATCH 6/6 arm: mx5: Add support for DENX M53EVK
Dear Marek Vasut,
Quoted 5 lines · Beno?t Th?baudeau
Dear Marek Vasut,
Quoted 2 lines · Marek Vasut
Dear Beno?t Th?baudeau,
Quoted 7 lines
Quoted 4 lines
[...]
Where did you get the information the board can only boot from NAND? Did I mistakenly present it somewhere like that please?
No, it's just suggested by the series, e.g. because imximage.cfg is NAND-specific.
This makes me wonder if the following should not be added to m53evk.h:
#define CONFIG_SPL_TARGET "u-boot-with-nand-spl.imx"
Best regards,
Beno?t
Fabio Estevam
reviewerPATCH 6/6 arm: mx5: Add support for DENX M53EVK
In reply to Marek VasutQuoted 2 lines · Marek Vasut
Add basic support for the DENX M53EVK board. Currently supported is the MMC, Ethernet, I2C.
What about NAND, SATA and USB ?
Quoted 11 lines
board/denx/m53evk/Makefile new file +8
-0@@ -0,0 +1,40 @@ +# +# DENX M53EVK +# Copyright (C) 2012-2013 Marek Vasut <[email protected]> +# +# 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.
board/denx/m53evk/Makefile line 18
Quoted 10 lines
board/denx/m53evk/Makefile new file +10
-0+# +# 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
You could just remove the adress paragraph. Same for other places in this patch.
board/denx/m53evk/Makefile line 28
Quoted 10 lines
board/denx/m53evk/Makefile new file +10
-0+u32 get_board_rev(void) +{ + struct iim_regs *iim = (struct iim_regs *)IMX_IIM_BASE; + struct fuse_bank *bank = &iim->bank[0]; + struct fuse_bank0_regs *fuse = + (struct fuse_bank0_regs *)bank->fuse_regs; + int rev = readl(&fuse->gp[6]); + + return (get_cpu_rev() & ~(0xF << 8)) | (rev & 0xF) << 8; +}
Can't you just remove this get_board_rev() from the board file and use the generic one instead?
board/denx/m53evk/Makefile line 37
Quoted 9 lines
board/denx/m53evk/Makefile new file +9
-0+/* + * Do not overwrite the console + * Use always serial for U-Boot console + */ +int overwrite_console(void) +{ + return 1; +} +
Is this really needed? It doesn't seem so, as you do not use splash screen.
board/denx/m53evk/Makefile
Quoted 7 lines
board/denx/m53evk/Makefile new file +7
-0+int board_init(void) +{ + gd->bd->bi_boot_params = PHYS_SDRAM_1 + 0x100; + + mxc_set_sata_internal_clock(); + + mxc_set_clock(0, 33, MXC_NFC_CLK);
Can you put a comment explaining which frequency you are setting MXC_NFC_CLK to ?
Quoted 3 lines
+/*
+ * U-Boot Commands
+ */
In this file you use multi-line style for single line comments.
You could simply do:
/* U-Boot Commands */
Quoted 1 line
+#define
CONFIG_SYS_MALLOC_LEN(10 * 1024 * 1024)
Why do you need such large size?
board/denx/m53evk/Makefile
Quoted 2 lines
board/denx/m53evk/Makefile new file +2
-0+#define CONFIG_SYS_MEMTEST_START 0x70010000 +#define CONFIG_SYS_MEMTEST_END 0x70020000
Are these values correct? You are only testing 64kB of the RAM.
PATCH 6/6 arm: mx5: Add support for DENX M53EVK
Dear Fabio Estevam,
[...]
Quoted 5 lines
Quoted 5 lines
board/denx/m53evk/m53evk.c new file +5
-0+u32 get_board_rev (void ) +{ +struct iim_regs *iim = (struct iim_regs *)IMX_IIM_BASE ; +struct fuse_bank *bank = &iim->bank[0 ]; + struct fuse_bank0_regs *fuse =
Quoted 8 lines
board/denx/m53evk/m53evk.c line 74
Quoted 5 lines
board/denx/m53evk/m53evk.c new file +5
-0+ (struct fuse_bank0_regs *)bank->fuse_regs; +int rev =readl (&fuse->gp[6 ]); + +return (get_cpu_rev () & ~(0xF <<8 )) | (rev &0xF ) <<8 ; +}Can't you just remove this
get_board_rev()from the board file and use the generic one instead?
There's no such thing for mx5, only for mx6.
I fixed the rest, but I left the multiline comments. It's the same on m28 and I consider it more readable.
Best regards,
Marek Vasut
PATCH 1/6 imx: Align the imximage header and payload to multiples of 4k
In reply to Marek VasutDear Marek Vasut,
Quoted 61 lines · Marek Vasut
The MX53 ROM loads the data from NAND in multiples of pages and supports maximum page size of 4k. Thus, align the image and header to 4k to be safe from ROM bugs.
Signed-off-by Marek Vasut <[email protected]>Cc Albert ARIBAUD <[email protected]>Cc Beno?t Th?baudeau <[email protected]>Cc Fabio Estevam <[email protected]>Cc Scott Wood <[email protected]>Cc Stefano Babic <[email protected]>Cc Tom Rini <[email protected]>--- tools/imximage.c | 11 +++++++---- tools/imximage.h | 3 ++- 2 files changed, 9 insertions(+), 5 deletions(-)tools/imximage.c +7
-4@@ -518,11 +518,14 @@ static void imximage_set_header(void *ptr, struct stat *sbuf, int ifd,/* * ROM bug alert -* mx53 only loads 512 byte multiples. -* The remaining fraction of a block bytes would -* not be loaded. +* +* MX53 only loads 512 byte multiples in case of SD boot. +* MX53 only loads NAND page multiples in case of NAND boot and +* supports up to 4096 byte large pages, thus align to 4096. +* +* The remaining fraction of a block bytes would not be loaded! */ - *header_size_ptr =ROUND (sbuf->st_size + imxhdr->flash_offset,512 ); + *header_size_ptr =ROUND (sbuf->st_size + imxhdr->flash_offset,4096 ); }int
imximage_check_params(struct mkimage_params *params)tools/imximage.h +2
-2@@ -151,13 +151,14 @@ typedef struct {dcd_v2_t dcd_table; } imx_header_v2_t; +/* The header must be aligned to 4k on MX53 for NAND boot */ struct imx_header {union {imx_header_v1_t hdr_v1;imx_header_v2_t hdr_v2; } header;uint32_t flash_offset; -}; +} __attribute__((aligned (4096 ))); typedef void (*set_dcd_val_t)(struct imx_header *imxhdr, char *name, int lineno, --
Quoted 1 line
1.7.11.7
Best regards,
Beno?t