The memory controller could already be enabled, when spd_sdram() is
called. This could be the case for example, when the SDRAM is initialized
by the JTAG debugger.
The "sync" after the register access via the accessor function is
still needed, because the macro uses the sync before the real write
is done. So until not all accesses are converted to using accessor
functions, this sync still needs to be made "manually" here.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Reinhard Arlt <reinhard.arlt@esd.eu>
Acked-by: Dave Liu <daveliu@freescale.com>
Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
Signed-off-by: Peter Tyser <ptyser@xes-inc.com>
Reviewed-by: Ira W. Snyder <iws@ovro.caltech.edu>
Tested-by: Ira W. Snyder <iws@ovro.caltech.edu>
Acked-by: Kim Phillips <kim.phillips@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Update 83xx architecture's CONFIG_ECC_INIT_VIA_DDRC references to
CONFIG_ECC_INIT_VIA_DDRCONTROLLER, which other Freescale architectures
use
Signed-off-by: Peter Tyser <ptyser@xes-inc.com>
Acked-by: Kim Phillips <kim.phillips@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Also update dmacpy()'s argument order to match memcpy's and use
phys_addr_t/phy_size_t for address/size arguments
Signed-off-by: Peter Tyser <ptyser@xes-inc.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
The routine assumed 4 bank SDRAMs, enhance to configure for 4 or 8
bank SDRAMs.
Signed-off-by: Gerald Van Baren <vanbaren@cideas.com>
Acked-by: Dave Liu <daveliu@freescale.com>
Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
1. RD_TO_PRE missed to add the AL, and need min 2 clocks for
tRTP according to DDR2 JEDEC spec.
2. WRTORD - tWTR need min 2 clocks according to DDR2 JEDEC spec.
3. add the support of DDR2-533,667,800 DIMMs
4. cpo
5. make the AL to min to gain better performance.
The Micron MT9HTF6472CHY-667D1 DIMMs test passed on
MPC837xEMDS platform at 266MHz/333MHz/400MHz data rate.
items 1, 2 and 5:
Acked-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se>
Reported-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se>
Signed-off-by: Dave Liu <daveliu@freescale.com>
Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
The spd_dram code shifts the base address, then masks 20 bits, but
forgets to shift the base address back. Fix this by just masking the
base address correctly.
Found this bug while trying to relocate a DDR memory at the base != 0.
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
This commit gets rid of a huge amount of silly white-space issues.
Especially, all sequences of SPACEs followed by TAB characters get
removed (unless they appear in print statements).
Also remove all embedded "vim:" and "vi:" statements which hide
indentation problems.
Signed-off-by: Wolfgang Denk <wd@denx.de>
32-bit wide ECC memory modules report 40-bit width.
Changed the DIMM data bus width test to 'less than 64' instead of 'equal 32'.
Signed-off-by: Lee Nipper <lee.nipper@freescale.com>
Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
delete ddr_clk and use mem_clk instead. Rename other ddr_*_clk to
mem_*_clk for consistency's sake.
Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
The DDR doesn't work on the 266MHz data rate,
the patch fix the bug.
Signed-off-by: Dave Liu <daveliu@freescale.com>
Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
The MPC837xEMDS board support:
* DDR2 400MHz hardcoded and SPD init
* Local bus NOR Flash
* I2C, UART, MII and RTC
* eTSEC RGMII
* PCI host
Signed-off-by: Dave Liu <daveliu@freescale.com>
introduced in the implement board_add_ram_info patch as I was cleaning out the
magic numbers. sorry.
Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
add board_add_ram_info, to make memory diagnostic output more
consistent. u-boot banner output now looks like:
DRAM: 256 MB (DDR1, 64-bit, ECC on)
and for boards with SDRAM on the local bus, a line such as this is
added:
SDRAM: 64 MB (local bus)
also replaced some magic numbers with their equivalent define names.
Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
Define CFG_DDR_SDRAM_CLK_CNTL for the MPC8349E-mITX and MPC8349E-mITX-GP.
This allows ddr->sdram_clk_cntl to be properly initialized. This is necessary
on some ITX boards, notably those with a revision 3.1 CPU.
Also change spd_sdram() in cpu/mpc83xx/spd_sdram.c to not write anything into
ddr->sdram_clk_cntl if CFG_DDR_SDRAM_CLK_CNTL is not defined.
Signed-off-by: Timur Tabi <timur@freescale.com>
Acked-by: Michael Benedict <MBenedict@twacs.com>
Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
There are a few bugs in the cpu/mpc83xx/spd_sdram.c
the first bug is that the picos_to_clk routine introduces a huge
rounding error in 83xx.
the second bug is that the mode register write recovery field is
tWR-1, not tWR >> 1.
The code supply fixed and SPD initialization for MPC83xx DDR2 Controller.
it pass DDR/DDR2 compliance tests.
Signed-off-by: Xie Xiaobo <X.Xie@freescale.com>
I've redone the SBC8349 support to match git-current, which
incorporates all the MPC834x updates from Freescale since the 1.1.6
release, including the DDR changes.
I've kept all the SBC8349 files as parallel as possible to the
MPC8349EMDS ones for ease of maintenance and to allow for easy
inspection of what was changed to support this board. Hence the SBC8349
U-Boot has FDT support and everything else that the MPC8349EMDS has.
Fortunately the Freescale updates added support for boards using CS0,
but I had to change spd_sdram.c to allow for board specific settings for
the sdram_clk_cntl (it is/was hard coded to zero, and that remains the
default if the board doesn't specify a value.)
Hopefully this should be mergeable as-is and require no whitespace
cleanups or similar, but if something doesn't measure up then let me
know and I'll fix it.
Thanks,
Paul.
Replace all instances of CFG_IMMRBAR with CFG_IMMR, so that the 83xx
tree matches the other 8xxx trees.
Signed-off-by: Timur Tabi <timur@freescale.com>
The 834x rev1.x silicon has one CPU5 errata.
The issue is when the data cache locked with
HID0[DLOCK], the dcbz instruction looks like no-op inst.
The right behavior of the data cache is when the data cache
Locked with HID0[DLOCK], the dcbz instruction allocates
new tags in cache.
The 834x rev3.0 and later and 8360 have not this bug inside.
So, when 834x rev3.0/8360 are working with ECC, the dcbz
instruction will corrupt the stack in cache, the processor will
checkstop reset.
However, the 834x rev1.x can work with ECC with these code,
because the sillicon has this cache bug. The dcbz will not
corrupt the stack in cache.
Really, it is the fault code running on fault sillicon.
This patch fix the incorrect dcbz operation. Instead of
CPU FP writing to initialise the ECC.
CHANGELOG:
* Fix the incorrect dcbz operation instead of CPU FP
writing to initialise the ECC memory. Otherwise, it
will corrupt the stack in cache, The processor will checkstop
reset.
Signed-off-by: Dave Liu <daveliu@freescale.com>
PREREQUISITE PATCHES:
* This patch can only be applied after the following patches have been applied:
1) DNX#2006090742000024 "Add support for multiple I2C buses"
2) DNX#2006090742000033 "Multi-bus I2C implementation of MPC834x"
3) DNX#2006091242000041 "Additional MPC8349 support for multibus i2c"
4) DNX#2006091242000078 "Add support for variable flash memory sizes on 83xx systems"
5) DNX#2006091242000069 "Add support for Errata DDR6 on MPC 834x systems"
CHANGELOG:
* Add support for the Freescale MPC8349E-mITX reference design platform.
The second TSEC (Vitesse 7385 switch) is not supported at this time.
Signed-off-by: Timur Tabi <timur@freescale.com>
CHANGELOG:
* Errata DDR6, which affects all current MPC 834x processors, lists changes
required to maintain compatibility with various types of DDR memory. This
patch implements those changes.
Signed-off-by: Timur Tabi <timur@freescale.com>
Incorporated the common unified variable names and the changes in preparation
for releasing mpc8360 patches.
Signed-off-by: Dave Liu <daveliu@freescale.com>
