[U-Boot-Users] "2.mach_types":(1 of 1) [PATCH][ARM] Create mach-types for various at91 boards
Author: Ulf Samuelsson ulf@atmel.com Date: 2007-03-27
Subject: "2.mach_types":(1 of 1) [PATCH][ARM] Create mach-types for various at91 boards
CHANGELOG: [PATCH][ARM] Create mach-types for various at91 boards at91rm9200df at91sam9260 at91sam9261 at91sam9263
Patch generated from files: include_asm-arm_mach-types.h.patch
Signed-off-by: Ulf Samuelsson --------------------------------------------------------------------------------------------------------------------------------- diff -urN u-boot-1.2.0/include/asm-arm/mach-types.h u-boot-1.2.0-atmel/include/asm-arm/mach-types.h --- u-boot-1.2.0/include/asm-arm/mach-types.h 2007-01-07 00:13:11.000000000 +0100 +++ u-boot-1.2.0-atmel/include/asm-arm/mach-types.h 2007-03-24 20:07:34.000000000 +0100 @@ -736,7 +736,11 @@ #define MACH_TYPE_LN2410SBC 725 #define MACH_TYPE_CB3RUFC 726 #define MACH_TYPE_MP2USB 727 +#define MACH_TYPE_AT91SAM9261EK 848 #define MACH_TYPE_PDNB3 1002 +#define MACH_TYPE_AT91SAM9260EK 1099 +#define MACH_TYPE_AT91RM9200DF 1119 +#define MACH_TYPE_AT91SAM9263EK 1202
#ifdef CONFIG_ARCH_EBSA110 # ifdef machine_arch_type @@ -9402,6 +9406,66 @@ # define machine_is_mp2usb() (0) #endif
+#ifdef CONFIG_MACH_AT91SAM9261EK +# ifdef machine_arch_type +# undef machine_arch_type +# define machine_arch_type __machine_arch_type +# else +# define machine_arch_type MACH_TYPE_AT91SAM9261EK +# endif +# define machine_is_at91sam9261ek() (machine_arch_type == MACH_TYPE_AT91SAM9261EK) +#else +# define machine_is_at91sam9261ek() (0) +#endif + +#ifdef CONFIG_MACH_AT91SAM9260EK +# ifdef machine_arch_type +# undef machine_arch_type +# define machine_arch_type __machine_arch_type +# else +# define machine_arch_type MACH_TYPE_AT91SAM9260EK +# endif +# define machine_is_at91sam9260ek() (machine_arch_type == MACH_TYPE_AT91SAM9260EK) +#else +# define machine_is_at91sam9260ek() (0) +#endif + +#ifdef CONFIG_MACH_AT91SAM9263EK +# ifdef machine_arch_type +# undef machine_arch_type +# define machine_arch_type __machine_arch_type +# else +# define machine_arch_type MACH_TYPE_AT91SAM9263EK +# endif +# define machine_is_at91sam9263ek() (machine_arch_type == MACH_TYPE_AT91SAM9263EK) +#else +# define machine_is_at91sam9263ek() (0) +#endif + +#ifdef CONFIG_MACH_AT91RM9200DF +# ifdef machine_arch_type +# undef machine_arch_type +# define machine_arch_type __machine_arch_type +# else +# define machine_arch_type MACH_TYPE_AT91RM9200DF +# endif +# define machine_is_at91rm9200df() (machine_arch_type == MACH_TYPE_AT91RM9200DF) +#else +# define machine_is_at91rm9200df() (0) +#endif + +#ifdef CONFIG_MACH_AT91SAM9263EK +# ifdef machine_arch_type +# undef machine_arch_type +# define machine_arch_type __machine_arch_type +# else +# define machine_arch_type MACH_TYPE_AT91SAM9263EK +# endif +# define machine_is_at91sam9263ek() (machine_arch_type == MACH_TYPE_AT91SAM9263EK) +#else +# define machine_is_at91sam9263ek() (0) +#endif + /* * These have not yet been registered */
In message 460942BA.3010609@atmel.com you wrote:
Subject: "2.mach_types":(1 of 1) [PATCH][ARM] Create mach-types for various at91 boards
CHANGELOG: [PATCH][ARM] Create mach-types for various at91 boards at91rm9200df at91sam9260 at91sam9261 at91sam9263
I send a NAK to this patch. Instead of updating individual entries, please update the whole file.
Best regards,
Wolfgang Denk
Wolfgang Denk skrev:
In message 460942BA.3010609@atmel.com you wrote:
Subject: "2.mach_types":(1 of 1) [PATCH][ARM] Create mach-types for various at91 boards
CHANGELOG: [PATCH][ARM] Create mach-types for various at91 boards at91rm9200df at91sam9260 at91sam9261 at91sam9263
I send a NAK to this patch. Instead of updating individual entries, please update the whole file.
Best regards,
Wolfgang Denk
You mean that I need to supply the complete board patch for the board? Intend to but each board will be too large for one file anyway.
In message 4609495C.6020802@atmel.com you wrote:
I send a NAK to this patch. Instead of updating individual entries, please update the whole file.
...
You mean that I need to supply the complete board patch for the board? Intend to but each board will be too large for one file anyway.
No. I mean that instead of updating just single machine type entries you should install a complete new versiosn of the mach-types file.
Best regards,
Wolfgang Denk
On Tue, Mar 27, 2007 at 06:38:21PM +0200, Wolfgang Denk wrote:
I send a NAK to this patch. Instead of updating individual entries, please update the whole file.
Sending again, this time gziped to fit into mailing list size limit. I almost forgot how big this file is...
Signed-off-by: Ladislav Michl ladis@linux-mips.org
Best regards, ladis
In message 20070523152851.GA25561@michl.2n.cz you wrote:
Sending again, this time gziped to fit into mailing list size limit. I almost forgot how big this file is...
Signed-off-by: Ladislav Michl ladis@linux-mips.org
Acked-by: Wolfgang Denk wd@denx.de
Content-Disposition: attachment; filename="mach-type.diff.gz"
Best regards,
Wolfgang Denk
On 3/27/07, Ulf Samuelsson ulf@atmel.com wrote:
diff -urN u-boot-1.2.0/include/asm-arm/mach-types.h u-boot-1.2.0-atmel/include/asm-arm/mach-types.h --- u-boot-1.2.0/include/asm-arm/mach-types.h 2007-01-07 00:13:11.000000000 +0100 +++ u-boot-1.2.0-atmel/include/asm-arm/mach-types.h 2007-03-24 20:07:34.000000000 +0100 @@ -736,7 +736,11 @@ #define MACH_TYPE_LN2410SBC 725 #define MACH_TYPE_CB3RUFC 726 #define MACH_TYPE_MP2USB 727 +#define MACH_TYPE_AT91SAM9261EK 848 #define MACH_TYPE_PDNB3 1002 +#define MACH_TYPE_AT91SAM9260EK 1099 +#define MACH_TYPE_AT91RM9200DF 1119 +#define MACH_TYPE_AT91SAM9263EK 1202
#ifdef CONFIG_ARCH_EBSA110 # ifdef machine_arch_type @@ -9402,6 +9406,66 @@ # define machine_is_mp2usb() (0) #endif
+#ifdef CONFIG_MACH_AT91SAM9261EK +# ifdef machine_arch_type +# undef machine_arch_type +# define machine_arch_type __machine_arch_type +# else +# define machine_arch_type MACH_TYPE_AT91SAM9261EK +# endif +# define machine_is_at91sam9261ek() (machine_arch_type == MACH_TYPE_AT91SAM9261EK) +#else +# define machine_is_at91sam9261ek() (0) +#endif
(Note: I realize this is a comment on the entire file, and your change just matches the existing code, but I'm going to bring it up anyway...)
I'm confused by this construct. What is the reason for the large #ifdef block? As I read it, there are two conditions for each board; 1. #ifdef CONFIG_MACH_XXXX 2. #ifdef machine_arch_type
Regardless of the result of condition 2, 'machine_arch_type' is always defined to something. That something either matches the current board, or it does not, and therefore the test in the first definition of machine_is_XXXXX() should always evaluate correctly.
To me, it seems like quite a verbose construct when something simpler could be used. For example:
#ifdef machine_arch_type # undef machine_arch_type # define machine_arch_type __machine_arch_type #else # ifdef CONFIG_MACH_ARMBOARD1 # define machine_arch_type MACH_TYPE_ARMBOARD1 # elif CONFIG_MACH_ARMBOARD2 # define machine_arch_type MACH_TYPE_ARMBOARD1 ...etc for all boards... # endif #endif
#define machine_is_board1() (machine_arch_type == MACH_TYPE_BOARD1) #define machine_is_board2() (machine_arch_type == MACH_TYPE_BOARD2) ...etc
Am I missing something?
Cheers, g.
In message 528646bc0703270956p682a183eu6af0545e33de8dd7@mail.gmail.com you wrote:
I'm confused by this construct. What is the reason for the large #ifdef block? As I read it, there are two conditions for each board;
You may ask this on the ARM linux mailing list. make sure to put your asbestos on...
Am I missing something?
Maybe - this whole file is a verbatim copy from the corresponding Linux kernel file.
And updating single entries just makes no sense. Instead, the recent version of the whole file should be copied.
And then we'll selentrly swallow that bitter pill of adding tons of unused defines just to get a few binary numbers...
Best regards,
Wolfgang Denk
On Tue, Mar 27, 2007 at 07:40:44PM +0200, Wolfgang Denk wrote:
Maybe - this whole file is a verbatim copy from the corresponding Linux kernel file.
And updating single entries just makes no sense. Instead, the recent version of the whole file should be copied.
And then we'll selentrly swallow that bitter pill of adding tons of unused defines just to get a few binary numbers...
Well, compared with
rsc@isonoe:~/git/u-boot$ find . -type f | grep -v .git | xargs wcgrep -i "#define" | wc -l 126281
it seems to fit the u-boot design very well ;)
Robert
participants (5)
-
Grant Likely -
Ladislav Michl -
Robert Schwebel -
Ulf Samuelsson -
Wolfgang Denk