[U-Boot] [PATCH v2] Noisily disable the legacy NAND subsystem.

Legacy NAND is marked for feature removal after April 2009 (i.e. this upcoming release). There are still several boards that reference it (though many do so only for disk-on-chip support which has been silently disabled for a while now). These boards will now fail to build with #error, though the code is still there if the user removes #error.
The plan is to remove the code outright in the next release, along with any board code that refers to it (such as board/esd/common/auto_update.c).
Also, remove the legacy NAND API description from README.nand.
Signed-off-by: Scott Wood scottwood@freescale.com --- Now with doc updates, restoration of a missing changelog line, and clarification of the error message in cmd_doc.c.
Applied to u-boot-nand-flash.
common/cmd_doc.c | 4 + doc/README.nand | 109 ++++----------------------------- drivers/mtd/nand_legacy/nand_legacy.c | 3 + 3 files changed, 20 insertions(+), 96 deletions(-)
diff --git a/common/cmd_doc.c b/common/cmd_doc.c index e2d4a42..3385c67 100644 --- a/common/cmd_doc.c +++ b/common/cmd_doc.c @@ -14,6 +14,10 @@ #include <linux/mtd/nftl.h> #include <linux/mtd/doc2000.h>
+#error This code is broken and will be removed outright in the next release. +#error If you need diskonchip support, please update the Linux driver in +#error drivers/mtd/nand/diskonchip.c to work with u-boot. + /* * ! BROKEN ! * diff --git a/doc/README.nand b/doc/README.nand index fc62f92..bb72289 100644 --- a/doc/README.nand +++ b/doc/README.nand @@ -98,83 +98,10 @@ Configuration Options: CONFIG_SYS_MAX_NAND_DEVICE The maximum number of NAND devices you want to support.
-NAND Interface: - - #define NAND_WAIT_READY(nand) - Wait until the NAND flash is ready. Typically this would be a - loop waiting for the READY/BUSY line from the flash to indicate it - it is ready. - - #define WRITE_NAND_COMMAND(d, adr) - Write the command byte `d' to the flash at `adr' with the - CLE (command latch enable) line true. If your board uses writes to - different addresses to control CLE and ALE, you can modify `adr' - to be the appropriate address here. If your board uses I/O registers - to control them, it is probably better to let NAND_CTL_SETCLE() - and company do it. - - #define WRITE_NAND_ADDRESS(d, adr) - Write the address byte `d' to the flash at `adr' with the - ALE (address latch enable) line true. If your board uses writes to - different addresses to control CLE and ALE, you can modify `adr' - to be the appropriate address here. If your board uses I/O registers - to control them, it is probably better to let NAND_CTL_SETALE() - and company do it. - - #define WRITE_NAND(d, adr) - Write the data byte `d' to the flash at `adr' with the - ALE and CLE lines false. If your board uses writes to - different addresses to control CLE and ALE, you can modify `adr' - to be the appropriate address here. If your board uses I/O registers - to control them, it is probably better to let NAND_CTL_CLRALE() - and company do it. - - #define READ_NAND(adr) - Read a data byte from the flash at `adr' with the - ALE and CLE lines false. If your board uses reads from - different addresses to control CLE and ALE, you can modify `adr' - to be the appropriate address here. If your board uses I/O registers - to control them, it is probably better to let NAND_CTL_CLRALE() - and company do it. - - #define NAND_DISABLE_CE(nand) - Set CE (Chip Enable) low to enable the NAND flash. - - #define NAND_ENABLE_CE(nand) - Set CE (Chip Enable) high to disable the NAND flash. - - #define NAND_CTL_CLRALE(nandptr) - Set ALE (address latch enable) low. If ALE control is handled by - WRITE_NAND_ADDRESS() this can be empty. - - #define NAND_CTL_SETALE(nandptr) - Set ALE (address latch enable) high. If ALE control is handled by - WRITE_NAND_ADDRESS() this can be empty. - - #define NAND_CTL_CLRCLE(nandptr) - Set CLE (command latch enable) low. If CLE control is handled by - WRITE_NAND_ADDRESS() this can be empty. - - #define NAND_CTL_SETCLE(nandptr) - Set CLE (command latch enable) high. If CLE control is handled by - WRITE_NAND_ADDRESS() this can be empty. - -More Definitions: - - These definitions are needed in the board configuration for now, but - may really belong in a header file. - TODO: Figure which ones are truly configuration settings and rename - them to CONFIG_SYS_NAND_... and move the rest somewhere appropriate. - - #define SECTORSIZE 512 - #define ADDR_COLUMN 1 - #define ADDR_PAGE 2 - #define ADDR_COLUMN_PAGE 3 - #define NAND_ChipID_UNKNOWN 0x00 - #define NAND_MAX_FLOORS 1 - #define CONFIG_SYS_NAND_MAX_CHIPS 1 - - #define CONFIG_SYS_DAVINCI_BROKEN_ECC + CONFIG_SYS_NAND_MAX_CHIPS + The maximum number of NAND chips per device to be supported. + + CONFIG_SYS_DAVINCI_BROKEN_ECC Versions of U-Boot <= 1.3.3 and Montavista Linux kernels generated bogus ECCs on large-page NAND. Both large and small page NAND ECCs were incompatible with the Linux davinci git tree (since @@ -186,27 +113,17 @@ More Definitions: NOTE: =====
-We now use a complete rewrite of the NAND code based on what is in -2.6.12 Linux kernel. - -The old NAND handling code has been re-factored and is now confined -to only board-specific files and - unfortunately - to the DoC code -(see below). A new configuration variable has been introduced: -CONFIG_NAND_LEGACY, which has to be defined in the board config file if -that board uses legacy code. - -The necessary changes have been made to all affected boards, and no -build breakage has been introduced, except for NETTA and NETTA_ISDN -targets from MAKEALL. This is due to the fact that these two boards -use JFFS, which has been adopted to use the new NAND, and at the same -time use NAND in legacy mode. The breakage will disappear when the -board-specific code is changed to the new NAND. +The current NAND implementation is based on what is in recent +Linux kernels. The old legacy implementation has been disabled, +and will be removed soon.
-As mentioned above, the legacy code is still used by the DoC subsystem. -The consequence of this is that the legacy NAND can't be removed from -the tree until the DoC is ported to use the new NAND support (or boards -with DoC will break). +If you have board code which used CONFIG_NAND_LEGACY, you'll need +to convert to the current NAND interface for it to continue to work.
+The Disk On Chip driver is currently broken and has been for some time. +There is a driver in drivers/mtd/nand, taken from Linux, that works with +the current NAND system but has not yet been adapted to the u-boot +environment.
Additional improvements to the NAND subsystem by Guido Classen, 10-10-2006
diff --git a/drivers/mtd/nand_legacy/nand_legacy.c b/drivers/mtd/nand_legacy/nand_legacy.c index 441780a..d9ae9c7 100644 --- a/drivers/mtd/nand_legacy/nand_legacy.c +++ b/drivers/mtd/nand_legacy/nand_legacy.c @@ -18,6 +18,9 @@ #include <linux/mtd/nand_ids.h> #include <jffs2/jffs2.h>
+#error Legacy NAND is deprecated. Please convert to the current NAND interface. +#error This code will be removed outright in the next release. + #ifdef CONFIG_OMAP1510 void archflashwp(void *archdata, int wp); #endif

On 16:41 Wed 01 Apr , Scott Wood wrote:
Legacy NAND is marked for feature removal after April 2009 (i.e. this upcoming release). There are still several boards that reference it (though many do so only for disk-on-chip support which has been silently disabled for a while now). These boards will now fail to build with #error, though the code is still there if the user removes #error.
The plan is to remove the code outright in the next release, along with any board code that refers to it (such as board/esd/common/auto_update.c).
Also, remove the legacy NAND API description from README.nand.
Signed-off-by: Scott Wood scottwood@freescale.com
Now with doc updates, restoration of a missing changelog line, and clarification of the error message in cmd_doc.c.
Applied to u-boot-nand-flash.
Scoot you brake the rm9200dk
could you convert your #error in #warning for this release as it's said in the commit message it will be remove for the next release
it will be nice to have this kind of #error at the beginning to the merge to allow people to fix it
Best Regards, J.

Dear Jean-Christophe PLAGNIOL-VILLARD,
In message 20090404163654.GE32409@game.jcrosoft.org you wrote:
Scoot you brake the rm9200dk
No, he did not. He just made the problem clearly visible - the code was has been broken before (just silently), if I understand correctly.
could you convert your #error in #warning for this release as it's said in the commit message it will be remove for the next release
I think we should leave this an #error, so people are really forced to react.
it will be nice to have this kind of #error at the beginning to the merge to allow people to fix it
Well, we are on the very first day of 10 week long bux fixing period. Can you imagine any better point in time to check this in? I cannot.
So please don't complain, either help fixing the broken code or remove the config option from the affected boards.
Best regards,
Wolfgang Denk

On Sat, 4 Apr 2009, Wolfgang Denk wrote:
Dear Jean-Christophe PLAGNIOL-VILLARD,
In message 20090404163654.GE32409@game.jcrosoft.org you wrote:
Scoot you brake the rm9200dk
No, he did not. He just made the problem clearly visible - the code was has been broken before (just silently), if I understand correctly.
could you convert your #error in #warning for this release as it's said in the commit message it will be remove for the next
release
I think we should leave this an #error, so people are really forced to react.
Second that. That ancient code should've long died.
--- ****************************************************************** * KSI@home KOI8 Net < > The impossible we do immediately. * * Las Vegas NV, USA < > Miracles require 24-hour notice. * ******************************************************************
participants (4)
-
Jean-Christophe PLAGNIOL-VILLARD
-
ksi@koi8.net
-
Scott Wood
-
Wolfgang Denk