Stefan Roese [Sat, 1 Apr 2006 11:41:03 +0000 (13:41 +0200)]
* Changes/fixes for drivers/cfi_flash.c:
- Add Intel legacy lock/unlock support to common CFI driver
On some Intel flash's (e.g. Intel J3) legacy unlocking is
supported, meaning that unlocking of one sector will unlock
all sectors of this bank. Using this feature, unlocking
of all sectors upon startup (via env var "unlock=yes") will
get much faster.
- Fixed problem with multiple reads of envronment variable
"unlock" as pointed out by Reinhard Arlt & Anders Larsen.
- Removed unwanted linefeeds from "protect" command when
CFG_FLASH_PROTECTION is enabled.
- Changed p3p400 board to use CFG_FLASH_PROTECTION
Patch by Stefan Roese, 01 Apr 2006
* Changes/fixes for drivers/cfi_flash.c:
- Correctly handle the cases where CFG_HZ != 1000 (several
XScale-based boards)
- Fix the timeout calculation of buffered writes (off by a
factor of 1000)
Patch by Anders Larsen, 31 Mar 2006
Wolfgang Denk [Mon, 13 Mar 2006 11:37:35 +0000 (12:37 +0100)]
Apply SoC concept to arm926ejs CPUs, i.e. move the SoC specific timer and
cpu_reset code from cpu/$(CPU) into the new cpu/$(CPU)/$(SOC) directories
Patch by Andreas Engel, 13 Mar 2006
Stefan Roese [Mon, 13 Mar 2006 10:16:36 +0000 (11:16 +0100)]
Change max size of uncompressed uImage's to 8MByte and add
CFG_BOOTM_LEN to adjust this setting.
As mentioned by Robin Getz on 2005-05-24 the size of uncompressed
uImages was restricted to 4MBytes. This default size is now
increased to 8Mbytes and can be overrided by setting CFG_BOOTM_LEN
in the board config file.
Wolfgang Denk [Sun, 12 Mar 2006 23:50:48 +0000 (00:50 +0100)]
Change the sequence of events in soft_i2c.c:send_ack() to keep from
incorrectly generating start/stop conditions on the bus.
Patch by Andrew Dyer, 26 Jul 2005
Wolfgang Denk [Sun, 12 Mar 2006 23:46:05 +0000 (00:46 +0100)]
Fix bug in [id]cache_status commands for MPC85xx processors;
should look at LSB of L1CSRn registers to determine if L1 cache is
enabled, not the MSB.
Patch by Murray Jensen, 19 Jul 2005
Wolfgang Denk [Sun, 12 Mar 2006 01:10:00 +0000 (02:10 +0100)]
Add ADI Blackfin support
- add support for Analog Devices Blackfin BF533 CPU
- add support for the ADI BF533 Stamp uClinux board
- add support for the ADI BF533 EZKit board
Patches by Richard Klingler, June 11th 2005:
Wolfgang Denk [Sun, 12 Mar 2006 00:48:55 +0000 (01:48 +0100)]
RPXlite configuration fixes
- Use correct flash sector size
- Use correct memory test end address
- Add support for bzip2 compression
- Various small fixes
Patch by Yuli Barcohen, 05 Jun 2005
Wolfgang Denk [Sun, 12 Mar 2006 00:45:44 +0000 (01:45 +0100)]
Memory configuration changes for ZPC.1900 board
- Fix SDRAM timing on both local bus and 60x bus
- Add support for second flash bank (SIMM)
- Change boot flash base
Patch by Yuli Barcohen, 05 Jun 2005
NAND finally working. Two bugs fixed:
* For READ_STATUS and READID commands always 8 bytes need to be read from
NDDB. Otherwise they stay there and get send to flash as the first data
word when writing.
* In nand_base.c the oob variable is not reset so this->oob_buf is
overwritten what eventually screws up the bad block descriptor table.
Wolfgang Denk [Mon, 6 Mar 2006 10:25:22 +0000 (11:25 +0100)]
Merge the new NAND code (testing-NAND brach); see doc/README.nand
Rewrite of NAND code based on what is in 2.6.12 Linux kernel
Patch by Ladislav Michl, 29 Jun 2005
Re-factoring the legacy NAND code (legacy NAND now only in board-specific
code and in SoC code). Boards using the old way have CFG_NAND_LEGACY and
BOARDLIBS = drivers/nand_legacy/libnand_legacy.a added. Build breakage for
NETTA.ERR and NETTA_ISDN - will go away when the new NAND support is
implemented for these boards.
Lots of new stuff:
* Debug message can be turned on and off.
* Waiting for events now times out.
* Implemented RESET command.
* Added appropriate nand_bbt_descriptor and nand_oobinfo.
Remaining Problems:
* Read Status still behaves weird an returns invalid stuff sometimes.
* ECC Placement does not respect our scheme in nand_oobinfo.
Moved the waiting loop for "Read Data Request" RDDREQ into the
delta_cmdfunc function, because this bit is only set once after a command
is sent and this allows read functions to be called multiple times.