There were a number of shared files that were using
CONFIG_SYS_MPC85xx_DDR_ADDR, or CONFIG_SYS_MPC86xx_DDR_ADDR, and
several variants (DDR2, DDR3). A recent patchset added
85xx-specific ones to code which was used by 86xx systems.
After reviewing places where these constants were used, and
noting that the type definitions of the pointers assigned to
point to those addresses were the same, the cleanest approach
to fixing this problem was to unify the namespace for the
85xx, 83xx, and 86xx DDR address definitions.
This patch does:
s/CONFIG_SYS_MPC8.xx_DDR/CONFIG_SYS_MPC8xxx_DDR/g
All 85xx, 86xx, and 83xx have been built with this change.
Signed-off-by: Andy Fleming <afleming@freescale.com>
Tested-by: Andy Fleming <afleming@freescale.com>
Acked-by: Kim Phillips <kim.phillips@freescale.com>
Also drop a few files referring to no longer / not yet supported
boards.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Prafulla Wadaskar <prafulla@marvell.com>
Cc: Stefan Roese <sr@denx.de>
Cc: Kim Phillips <kim.phillips@freescale.com>
Cc: Andy Fleming <afleming@gmail.com>
Cc: Jason Jin <jason.jin@freescale.com>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Daniel Schwierzeck <daniel.schwierzeck@googlemail.com>
Acked-by: Stefano Babic <sbabic@denx.de>
Acked-by: Daniel Schwierzeck <daniel.schwierzeck@googlemail.com>
The code here was copied from the mpc8548cds support, and it
wasn't using the CONFIG_SYS_LBC_LCRR define, and was just
unconditionally setting the LCRR_EADC bit. Snooping with a
hardware debugger also showed we had LCRR_DBYP set, since we were
setting it based on a read of an uninitialized lcrr read via
clkdiv. Borrow from the code in the tqm85xx.c support to add
LBC frequency aware masking of these bits.
This change will correct reliability issues associated with trying
to use the 128MB of LBC 100MHz SDRAM on this board. Thanks to
Keith Savage for assistance in diagnosing the root cause of this.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Existing boards by default have an issue where the LBC SDRAM
SPD EEPROM and the DDR2 SDRAM SPD EEPROM both land at 0x51.
After the hardware modification listed in the README is made,
then the DDR2 SPD EEPROM appears at 0x53. So this implements
a board specific get_spd() by taking advantage of the existing
weak linkage, that 1st tries reading at 0x53 and then if that
fails, it falls back to the old 0x51.
Since the old dependency issue of "SPD implies no LBC SDRAM"
gets removed with the hardware errata fix, remove that restriction
in the code, so both LBC SDRAM and SPD can be selected.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Nothing to see here, just a relocation of the fixed ddr init
sequence to live in the actual ddr.c file itself.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Previously, SPD configuration of RAM was non functional on
this board. Now that the root cause is known (an i2c address
conflict), there is a simple end-user workaround - remove the
old slower local bus 128MB module and then SPD detection on the
main DDR2 memory module works fine.
We make the enablement of the LBC SDRAM support conditional on
being not SPD enabled. We can revisit this dependency as the
hardware workaround becomes available.
Turning off LBC SDRAM support revealed a couple implict dependencies
in the tlb/law code that always expected an LBC SDRAM address.
This has been tested with the default 256MB module, a 512MB
a 1GB and a 2GB, of varying speeds, and the SPD autoconfiguration
worked fine in all cases.
The default configuration remains to go with the hard coded
DDR config, so the default build will continue to work on boards
where people don't bother to read the docs. But the advantage
of going to the SPD config is that even the small default module
gets configured for CL3 instead of CL4.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
These were cloned from the mpc8548cds platform which has
a different memory layout (1/2 the size). Set the values
by comparing to the register file for the board used during
JTAG init sequence:
LSDMR1 0x2863B727 /* PCHALL */
LSDMR2 0x0863B727 /* NORMAL */
LSDMR3 0x1863B727 /* MRW */
LSDMR4 0x4063B727 /* RFEN */
This differs from what was there already in that the RFEN is
not bundled in all four steps implicitly, but issued once
as the final step.
The other difference seen when comparing vs. the register file init,
is that since the memory is split across /CS3 and /CS4, the dummy
writes need to go to 0xf000_0000 _and_ to 0xf400_0000.
We also rewrite the final LBC SDRAM inits as macros, as there is
no real need for them to be a local variable that is modified
on the fly at runtime.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
This board has an 8MB soldered on flash, and a 64MB SODIMM
flash module. Normally the board boots from the 8MB flash,
but the hardware can be configured for booting from the 64MB
flash as well by swapping CS0 and CS6. This can be handy
for recovery purposes, or for supporting u-boot and VxBoot
at the same time.
To support this in u-boot, we need to have different BR0/OR0
and BR6/OR6 settings in place for when the board is configured
in this way, and a different TEXT_BASE needs to be used due
to the larger sector size of the 64MB flash module.
We introduce the suffix _8M and _64M for the BR0/BR6 and the
OR0/OR6 values so it is clear which is being used to map what
specific device.
The larger sector size (512k) of the alternate flash needs
a larger malloc pool, otherwise you'll get failures when
running saveenv, so bump it up accordingly.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
The current situation has the 64MB user flash at an awkward
alignment; shifted back from 0xfc00_0000 by 8M, to leave an 8MB hole
for the soldered on boot flash @ EOM. But to switch to optionally
supporting booting off the 64MB flash, the 64MB will then be mapped
at the sane address of 0xfc00_0000.
This leads to awkward things when programming the 64MB flash prior
to transitioning to it -- i.e. even though the chip spans from
0xfb80_0000 to 0xff7f_ffff, you would have to program a u-boot image
into the two sectors from 0xfbf0_0000 --> 0xfbff_ffff so that it was
in the right place when JP12/SW2.8 were switched to make the 64MB on
/CS0. (i.e. the chip is only looking at the bits in mask 0x3ff_ffff)
We also have to have three TLB entries responsible for dealing with
mapping the 64MB flash due to this 8MB of misalignment.
In the end, there is address space from 0xec00_0000 to 0xefff_ffff
where we can map it, and then the transition from booting from one
config to the other will be a simple 0xec --> 0xfc mapping. Plus we
can toss out a TLB entry.
Note that TLB0 is kept at 64MB and not shrunk down to the 8MB boot
flash; this means we won't have to change it when the alternate
config uses the full 64MB for booting, in TLB0.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Fix:
sbc8548.c: In function 'local_bus_init':
sbc8548.c:80:7: warning: variable 'lbc_hz' set but not used [-Wunused-but-set-variable]
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
The top level Makefile does not do any recursion into subdirs when
cleaning, so these clean/distclean targets in random arch/board dirs
never get used. Punt them all.
MAKEALL didn't report any errors related to this that I could see.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Move fsl_ddr_get_spd into common mpc8xxx/ddr/main.c as most boards
pretty much do the same thing. The only variations are in how many
controllers or DIMMs per controller exist. To make this work we
standardize on the names of the SPD_EEPROM_ADDRESS defines based on the
use case of the board.
We allow boards to override get_spd to either do board specific fixups
to the SPD data or deal with any unique behavior of how the SPD eeproms
are wired up.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Every 85xx board implements fsl_ddr_get_mem_data_rate via get_ddr_freq()
and every 86xx board uses get_bus_freq(). If implement get_ddr_freq()
as a static inline to call get_bus_freq() we can remove
fsl_ddr_get_mem_data_rate altogether and just call get_ddr_freq()
directly.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Remove duplicated code in SBC8548 board and utilize the common
fsl_pcie_init_board(). We also now dynamically setup the LAWs for PCI
controllers based on which PCIe controllers are enabled.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Tested-by: Paul Gortmaker <paul.gortmaker@windriver.com>
sdram_init() is used to initialize sdram on the lbc. Rename it
accordingly.
Signed-off-by: Becky Bruce <beckyb@kernel.crashing.org>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Correct initdram to use phys_size_t to represent the size of
dram; instead of changing this all over the place, and correcting
all the other random errors I've noticed, create a
common initdram that is used by all non-corenet 85xx parts. Most
of the initdram() functions were identical, with 2 common differences:
1) DDR tlbs for the fixed_sdram case were set up in initdram() on
some boards, and were part of the tlb_table on others. I have
changed them all over to the initdram() method - we shouldn't
be accessing dram before this point so they don't need to be
done sooner, and this seems cleaner.
2) Parts that require the DDR11 erratum workaround had different
implementations - I have adopted the version from the Freescale
errata document. It also looks like some of the versions were
buggy, and, depending on timing, could have resulted in the
DDR controller being disabled. This seems bad.
The xpedite boards had a common/fsl_8xxx_ddr.c; with this
change only the 517 board uses this so I have moved the ddr code
into that board's directory in xpedite517x.c
Signed-off-by: Becky Bruce <beckyb@kernel.crashing.org>
Tested-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Now that we have serdes support for all 85xx/86xx/Pxxx chips we can
replace the is_fsl_pci_cfg() code with the is_serdes_configured().
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Before this commit, weak symbols were not overridden by non-weak symbols
found in archive libraries when linking with recent versions of
binutils. As stated in the System V ABI, "the link editor does not
extract archive members to resolve undefined weak symbols".
This commit changes all Makefiles to use partial linking (ld -r) instead
of creating library archives, which forces all symbols to participate in
linking, allowing non-weak symbols to override weak symbols as intended.
This approach is also used by Linux, from which the gmake function
cmd_link_o_target (defined in config.mk and used in all Makefiles) is
inspired.
The name of each former library archive is preserved except for
extensions which change from ".a" to ".o". This commit updates
references accordingly where needed, in particular in some linker
scripts.
This commit reveals board configurations that exclude some features but
include source files that depend these disabled features in the build,
resulting in undefined symbols. Known such cases include:
- disabling CMD_NET but not CMD_NFS;
- enabling CONFIG_OF_LIBFDT but not CONFIG_QE.
Signed-off-by: Sebastien Carlier <sebastien.carlier@gmail.com>
Clean up Makefile, and drop a lot of the config.mk files on the way.
We now also automatically pick all boards that are listed in
boards.cfg (and with all configurations), so we can drop the redundant
entries from MAKEALL to avoid building these twice.
Signed-off-by: Wolfgang Denk <wd@denx.de>
The change is currently needed to be able to remove the board
configuration scripting from the top level Makefile and replace it by
a simple, table driven script.
Moving this configuration setting into the "CONFIG_*" name space is
also desirable because it is needed if we ever should move forward to
a Kconfig driven configuration system.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Previously we used an alias the pci node to determine which node to
fixup or delete. Now we use the new fdt_node_offset_by_compat_reg to
find the node to update.
Additionally, we replace the code in each board with a single macro call
that makes assumes uniform naming and reduces duplication in this area.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Currently, 83xx, 86xx, and 85xx have a lot of duplicated code
dedicated to defining and manipulating the LBC registers. Merge
this into a single spot.
To do this, we have to decide on a common name for the data structure
that holds the lbc registers - it will now be known as fsl_lbc_t, and we
adopt a common name for the immap layouts that include the lbc - this was
previously known as either im_lbc or lbus; use the former.
In addition, create accessors for the BR/OR regs that use in/out_be32
and use those instead of the mismash of access methods currently in play.
I have done a successful ppc build all and tested a board or two from
each processor family.
Signed-off-by: Becky Bruce <beckyb@kernel.crashing.org>
Acked-by: Kim Phillips <kim.phillips@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
This reverts commit 70ed869ea5.
There isn't any need to modify the API for fsl_pci_init_port to pass the
status of host/agent(end-point) status. We can determine that
internally to fsl_pci_init_port. Revert the patch that makes the API
change.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Originally written by Jason Jin and Mingkai Hu for mpc8536.
When QorIQ based board is configured as a PCIe agent, then unlock/enable
inbound PCI configuration cycles and init a 4K inbound memory window;
so that a PCIe host can access the PCIe agents SDRAM at address 0x0
* Supported in fsl_pci_init_port() after adding pcie_ep as a param
* Revamped copyright in drivers/pci/fsl_pci_init.c
* Mods in 85xx based board specific pci init after this change
Signed-off-by: Vivek Mahajan <vivek.mahajan@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
By nature of being based off the MPC8548CDS board, this
board inherited an ENV_SIZE setting of 256k. But since
it has a smaller flash device (8MB soldered on), it has
a native sector size of 128k, and hence the ENV_SIZE was
causing 2 sectors to be used for the environment.
By removing the unused sector, we can push TEXT_BASE up
closer to the end of address space and reclaim that
sector for any other application. This also fixes the
mismatch between TEXT_BASE and MONITOR_LEN reported by
Kumar earlier.
Since this board also supports the ability to boot off
the 64MB SODIMM flash, this change is forward looking
with that in mind; i.e. the settings for MONITOR_LEN
and ENV_SIZE will work when the 512k sectors of the
SODIMM flash are used for alternate boot in the future.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
The PCI/PCI-e support for the sbc8548 was based on an earlier
version of what the MPC8548CDS board was using, and in its
current state it won't even compile. This re-syncs it to match
the latest codebase and makes use of the new shared PCI functions
to reduce board duplication.
It borrows from the MPC8568MDS, in that it pulls the PCI-e I/O
back to 0xe280_0000 (where PCI2 would be on MPC8548CDS), and
similarly it coalesces the PCI and PCI-e mem into one single TLB.
Both PCI-x and PCI-e have been tested with intel e1000 cards
under linux (with an accompanying dts change in place)
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
The size of the LB SDRAM on this board is 128MB, spanning CS3
and CS4. It was previously only being configured for 64MB on
CS3, since that was what the original codebase of the MPC8548CDS
had. In addition to setting up BR4/OR4, this also adds the TLB
entry for the second half of the SDRAM.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Sweep throught the board specific file and replace the various
register proddings with the equivalent I/O accessors.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
With only eTSEC1 and 2 being brought out to RJ-45 connectors, we
aren't interested in the eTSEC3/4 voltage hack on this board
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
The sbc8548 has a 64MB SODIMM flash module off of CS6 that
previously wasn't enumerated by u-boot. There were already
BR6/OR6 settings for it [used by cpu_init_f()] but there
was no TLB entry and it wasn't in the list of flash banks
reported to u-boot.
The location of the 64MB flash is "pulled back" 8MB from
a 64MB boundary, in order to allow address space for the
8MB boot flash that is at the end of 32 bit address space.
This means creating two 4MB TLB entries for the 8MB chunk,
and then expanding the original boot flash entry to 64MB
in order to cover the 8MB boot flash and the remainder
(56MB) of the user flash.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Fix the extra long lines to be consistent with u-boot coding style.
No functional change here.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
The get_clock_freq() comes from freescale/common/cadmus.c and is
only valid for the CDS based 85xx reference platforms. It would
be nice if we could read the 33 vs. 66MHz status somehow, but in
the meantime, tie it to CONFIG_SYS_CLK_FREQ like all the other
non-CDS boards do.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
There are a couple defines and PCI bridge quirks related to the PCI
backplane of the MPC8548CDS that have no meaning in the context of
the port to the sbc8548 board, so delete them.
Also, the form factor of the sbc8548 is a standalone board with a
single PCI-X and a single PCI-e slot. That pretty much guarantees
that it will never be a PCI agent itself, so the host/agent and root
complex/end node distinctions have been removed.
Similarly, since there is no physical connector mapping to PCI2, so
all references of PCI2 in the board support files have been removed
as well.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Create a board_eth_init to allow a place to hook in
the PCI ethernet init after all the eTSEC are up
and configured.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
The previous README.sbc8548 was pretty much content-free. Replace
it with something that actually gives the end user some relevant
hardware details, and also lists the u-boot configuration choices.
Also in the cosmetic department, fix the bogus line in the Makefile
that was carried over from the SBC8560 Makefile, and the typo in
the sbc8548.c copyright.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Refactor the code into a simple bitmask lookup table that determines if
a given PCI controller is enabled and if its in host/root-complex or
agent/end-point mode.
Each processor in the PQ3/MPC86xx family specified different encodings
for the cfg_host_agt[] and cfg_IO_ports[] boot strapping signals.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
There are really no differences between all the 85xx linker scripts so
we can just move to a single common one. Board code is still able to
override the common one if need be.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
For historic reasons we had defined some additional PLATFORM_CPPFLAGS
like:
PLATFORM_CPPFLAGS += -DCONFIG_E500=1
PLATFORM_CPPFLAGS += -DCONFIG_MPC85xx=1
PLATFORM_CPPFLAGS += -DCONFIG_MPC8548=1
However these are all captured in the config.h and thus redudant.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Every platform that calls fsl_pci_init calls fsl_pci_setup_inbound_windows
before it calls fsl_pci_init. There isn't any reason to just call it
from fsl_pci_init and simplify things a bit.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Every platform that calls fsl_pci_init calls pci_setup_indirect before
it calls fsl_pci_init. There isn't any reason to just call it from
fsl_pci_init and simplify things a bit.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Rename the pci header for FSL HW so we can move some prototypes
in there and stop doing explicit externs
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
A recent gcc added a new unaligned rodata section called '.rodata.str1.1',
which needs to be added the the linker script. Instead of just adding this
one section, we use a wildcard ".rodata*" to get all rodata linker section
gcc has now and might add in the future.
However, '*(.rodata*)' by itself will result in sub-optimal section
ordering. The sections will be sorted by object file, which causes extra
padding between the unaligned rodata.str.1.1 of one object file and the
aligned rodata of the next object file. This is easy to fix by using the
SORT_BY_ALIGNMENT command.
This patch has not be tested one most of the boards modified. Some boards
have a linker script that looks something like this:
*(.text)
. = ALIGN(16);
*(.rodata)
*(.rodata.str1.4)
*(.eh_frame)
I change this to:
*(.text)
. = ALIGN(16);
*(.eh_frame)
*(SORT_BY_ALIGNMENT(SORT_BY_NAME(.rodata*)))
This means the start of rodata will no longer be 16 bytes aligned.
However, the boundary between text and rodata/eh_frame is still aligned to
16 bytes, which is what I think the real purpose of the ALIGN call is.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
The PCI_REGION_MEMORY and PCI_REGION_MEM are a bit to similar and
can be confusing when reading the code.
Rename PCI_REGION_MEMORY to PCI_REGION_SYS_MEMORY to clarify its used
for system memory mapping purposes.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
On newer CPUs, 8536, 8572, and 8610, the CLKDIV field of LCRR is five bits
instead of four.
In order to avoid an ifdef, LCRR_CLKDIV is set to 0x1f on all systems. It
should be safe as the fifth bit was defined as reserved and set to 0.
Code that was using a hard coded 0x0f is changed to use LCRR_CLKDIV.
Signed-off-by: Trent Piepho <tpiepho@freescale.com>
Acked-by: Kumar Gala <galak@kernel.crashing.org>
Acked-by: Jon Loeliger <jdl@freescale.com>
Add define used to determine if PCI1 interface is in PCI or PCIX mode.
Convert users of the old PORDEVSR_PCI constant to use MPC85xx_PORDEVSR_PCI1
Signed-off-by: Peter Tyser <ptyser@xes-inc.com>
Signed-off-by: Andy Fleming <afleming@freescale.com>
All mpc8548-based boards should implement the suggested workaround
to CPU 2 errata. Without the workaround, its possible for the
8548's core to hang while executing a msync or mbar 0 instruction
and a snoopable transaction from an I/O master tagged to make
quick forward progress is present.
Signed-off-by: Peter Tyser <ptyser@xes-inc.com>
Acked-by: Andy Fleming <afleming@freescale.com>
The DDR controller of 8548/8544/8568/8572/8536 processors
have the ECC data init feature, and the new DDR code is
using the feature, and we don't need the way with DMA to
init memory any more.
Signed-off-by: Dave Liu <daveliu@freescale.com>
Acked-by: Andy Fleming <afleming@freescale.com>
