As explained by Wolfgang, historically PowerPC would do a number of
things to hand-optimize placement of the binary on NOR flash in order to
maximize utilization of very scarce resources. These days, we simply
aren't optimizing our binary layout for NOR flash placement and it's
quite likely this wasn't working as intended. Furthermore, this level
of optimization makes it difficult to have version_string be a global,
instead of a weak and overridden value, and so make more progress on
reproducible builds, which is a current concern.
Move to having PowerPC no longer store version_string in the early part
of text so that it might be part of the first page of NOR and instead
use the same declaration everyone else does.
Link: https://lore.kernel.org/r/96716.1629798400@gemini.denx.de/
Signed-off-by: Tom Rini <trini@konsulko.com>
This board has not been converted to CONFIG_DM_MMC by the deadline.
Remove it. As this is the last ARCH_MPC8569 board, remove that support
as well.
Cc: Priyanka Jain <priyanka.jain@nxp.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Priyanka Jain <priyanka.jain@nxp.com>
This is an init function so move it out of the common header. Avoid using
the typedef so that we don't have to include the global_data header file.
Also tidy up the function style in comments while we are here.
Signed-off-by: Simon Glass <sjg@chromium.org>
Rename the CONFIG_SECURE_BOOT name to CONFIG_NXP_ESBC to avoid
conflicts with UEFI secure boot.
Signed-off-by: Udit Agarwal <udit.agarwal@nxp.com>
Signed-off-by: Priyanka Jain <priyanka.jain@nxp.com>
-msingle-pic-base is a new gcc(from 4.6) option for ppc and
it reduces the size of my u-boot with about 4-5 KB.
While at it, add -fno-jump-tables too to save a
few more bytes.
e5500 core:
size u-boot.bef
text data bss dec hex filename
473043 23772 307104 803919 c444f u-boot.bef
size u-boot.aft
text data bss dec hex filename
453195 23772 307104 784071 bf6c7 u-boot.aft
e500 core:
size u-boot.bef
text data bss dec hex filename
292998 17868 24968 335834 51fda u-boot.bef
size u-boot.aft
text data bss dec hex filename
288002 17868 24968 330838 50c56 u-boot.aft
Signed-off-by: Joakim Tjernlund <joakim.tjernlund@infinera.com>
Reviewed-by: York Sun <york.sun@nxp.com>
When U-Boot started using SPDX tags we were among the early adopters and
there weren't a lot of other examples to borrow from. So we picked the
area of the file that usually had a full license text and replaced it
with an appropriate SPDX-License-Identifier: entry. Since then, the
Linux Kernel has adopted SPDX tags and they place it as the very first
line in a file (except where shebangs are used, then it's second line)
and with slightly different comment styles than us.
In part due to community overlap, in part due to better tag visibility
and in part for other minor reasons, switch over to that style.
This commit changes all instances where we have a single declared
license in the tag as both the before and after are identical in tag
contents. There's also a few places where I found we did not have a tag
and have introduced one.
Signed-off-by: Tom Rini <trini@konsulko.com>
Some platforms have very limited SRAM to run SPL code, so there may
not be the same amount space for a malloc pool before relocation in
the SPL stage as the normal U-Boot stage.
Make SPL and (the full) U-Boot stage use independent SYS_MALLOC_F_LEN,
so the size of pre-relocation malloc pool can be configured memory
space independently.
Signed-off-by: Andy Yan <andy.yan@rock-chips.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
[fixed up commit-message:]
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
For E6500 cores, L2 cache has been used as init_ram. L1 cache is a
write through cache on E6500.If lines are not locked in both L1 and
L2 caches, crashes are observed during secure boot. This patch locks/
unlocks both L1 and L2 cache to prevent the crash.
Signed-off-by: Ruchika Gupta <ruchika.gupta@nxp.com>
Reviewed-by: York Sun <york.sun@nxp.com>
We should use the symbolic names for the cache control bits.
Signed-off-by: Mark Marshall <Mark.Marshall@omicron.at>
Reviewed-by: Thomas Graziadei <thomas.graziadei@omicronenergy.com>
Reviewed-by: York Sun <york.sun@nxp.com>
For mpc85xx SoCs, the core begins execution from address 0xFFFFFFFC.
In non-secure boot scenario from NAND, this address will map to CPC
configured as SRAM. But in case of secure boot, this default address
always maps to IBR (Internal Boot ROM).
The IBR code requires that the bootloader(U-boot) must lie in 0 to 3.5G
address space i.e. 0x0 - 0xDFFFFFFF.
For secure boot target from NAND, the text base for SPL is kept same as
non-secure boot target i.e. 0xFFFx_xxxx but the SPL U-boot binary will
be copied to CPC configured as SRAM with address in 0-3.5G(0xBFFC_0000)
As a the virtual and physical address of CPC would be different. The
virtual address 0xFFFx_xxxx needs to be mapped to physical address
0xBFFx_xxxx.
Create a new PBI file to configure CPC as SRAM with address 0xBFFC0000
and update DCFG SCRTACH1 register with location of Header required for
secure boot.
The changes are similar to
commit 467a40dfe3
powerpc/mpc85xx: SECURE BOOT- NAND secure boot target for P3041
While P3041 has a 1MB CPC and does not require SPL. On T104x, CPC
is only 256K and thus SPL framework is used.
The changes are only applicable for SPL U-Boot running out of CPC SRAM
and not the next level U-Boot loaded on DDR.
Reviewed-by: Ruchika Gupta <ruchika.gupta@nxp.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Aneesh Bansal <aneesh.bansal@nxp.com>
Signed-off-by: Sumit Garg <sumit.garg@nxp.com>
Reviewed-by: York Sun <york.sun@nxp.com>
While enabling L2 cache, the value of L2PE (L2 cache parity/ECC
error checking enable) must not be changed while the L2 cache is
enabled.
So, L2PE must be set before enabling L2 cache.
Signed-off-by: Aneesh Bansal <aneesh.bansal@nxp.com>
Reviewed-by: York Sun <york.sun@nxp.com>
To enable DM on MPC85xx, we need pre-relocation malloc, which is
implemented in this patch.
We also make sure that the IVORs are always 4-aligned on e500 to prevent
alignment exceptions caused by code changes in start.S.
Signed-off-by: Mario Six <mario.six@gdsys.cc>
Cc: York Sun <york.sun@nxp.com>
Cc: Simon Glass <sjg@chromium.org>
Reviewed-by: York Sun <york.sun@nxp.com>
MPC85xx has been using locked L1 cache as init_ram. L1 cache is a write
through cache on E6500. L2 cache is enabled to to hold the data. This
patch locks/unlocks L2 cache to ensure no data cast out from L2 cache.
Signed-off-by: York Sun <yorksun@freescale.com>
Reported-by: Jeffery Zhu <Jefferry.Zhu@freescale.com>
Secure Boot Target is added for NAND for P3041.
For mpc85xx SoCs, the core begins execution from address 0xFFFFFFFC.
In case of secure boot, this default address maps to Boot ROM.
The Boot ROM code requires that the bootloader(U-boot) must lie
in 0 to 3.5G address space i.e. 0x0 - 0xDFFFFFFF.
In case of NAND Secure Boot, CONFIG_SYS_RAMBOOT is enabled and CPC is
configured as SRAM. U-Boot binary will be located on SRAM configured
at address 0xBFF00000.
In the U-Boot code, TLB entries are created to map the virtual address
0xFFF00000 to physical address 0xBFF00000 of CPC configured as SRAM.
Signed-off-by: Saksham Jain <saksham@freescale.com>
Signed-off-by: Ruchika Gupta <ruchika.gupta@freescale.com>
Signed-off-by: Aneesh Bansal <aneesh.bansal@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
Commit 96d2bb952b ("powerpc/mpc85xx: Don't relocate exception vectors")
simplified IVOR initialization a bit too much, failing to use the
post-relocation offset. This doesn't cause a problem with normal NOR
boot, in which both the pre-relocation and post-relocation addresses
are 64 KiB aligned. However, if TEXT_BASE is only 4 KiB aligned, such
as for NAND/SD/etc. boot on some targets, as well as the QEMU target,
the post-relocation address will not be the same in the lower 16 bits,
as reserve_uboot() ensures that the relocation address is always 64
KiB aligned even if the pre-relocation address was not.
Use the GOT to get the proper post-relocation offsets.
Fixes: 96d2bb952b ("powerpc/mpc85xx: Don't relocate exception vectors")
Signed-off-by: Scott Wood <scottwood@freescale.com>
Cc: Alexander Graf <agraf@suse.de>
Cc: Shaohui Xie <Shaohui.Xie@freescale.com>
Tested-by: Shaohui Xie <Shaohui.Xie@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
U-Boot does not have system calls (the services it exposes to
standalone commands use a different mechanism), so the syscall handler
is dead code. It's also broken code, as it assumes it is located at
0xc00 -- while even before the patch to stop relocating exception
vectors to 0, U-Boot had the syscall at 0x900.
The critical and machine check return paths are never called -- the
regular exception return path is used instead, which works because
xSRR0/1 have already been saved and can be restored via the regular
SRR0/1 (we don't care too much in U-Boot about taking a critical/mcheck
inside another exception prolog/epilog).
Also remove a few other small unused functions.
Signed-off-by: Scott Wood <scottwood@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
Booke does not require exception vectors to be located at address zero.
U-Boot was doing so anyway, simply because that's how it had been done
on other PPC. The downside of this is that once the OS is loaded to
address zero, the exception vectors have been overwritten -- which
makes it difficult to diagnose a crash that happens after that point.
The IVOR setup and trap entry code is simplified somewhat as a result.
Also, there is no longer a need to align individual exceptions on 0x100
byte boundaries.
Signed-off-by: Scott Wood <scottwood@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
baord_init_f takes one argument, boot_flag. It has not been used for
powerpc, until recently changing to use generic board architecture.
The boot flag is added as a return value from cpu_init_f().
Signed-off-by: York Sun <yorksun@freescale.com>
CC: Alexander Graf <agraf@suse.de>
For the QEMU machine type, we can plug in either e500v2, e500mc, e5500
or e6500 style cores into the system. U-boot has to work with all of them.
So avoid using HID1 which is not available on e500mc systems to make sure
we don't trap on it.
Signed-off-by: Alexander Graf <agraf@suse.de>
Current SPL code base has BSS section placed after reset_vector. This means
they have to relocate to use the global variables. This put an implicit
requirement of having SPL size = Memory/2.
To avoid relocation:
- Move bss_section within SPL range
- Modify relocate_code()
Signed-off-by: Prabhakar Kushwaha <prabhakar@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
LAW_EN is only defined if CONFIG_SYS_CCSRBAR_DEFAULT is not equal to
CONFIG_SYS_CCSRBAR_PHYS. in SPL framework CCSRBAR is not relocated hence
both are same. This cause compilation error.
So LAW_EN define outside of configs
Signed-off-by: Prabhakar Kushwaha <prabhakar@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
Changes:
1. L2 cache is being invalidated by Boot ROM code for e6500 core.
So removing the invalidation from start.S
2. Clear the LAW and corresponding configuration for CPC. Boot ROM
code uses it as hosekeeping area.
3. For Secure boot, CPC is configured as SRAM and used as house
keeping area. This configuration is to be disabled once in uboot.
Earlier this disabling of CPC as SRAM was happening in cpu_init_r.
As a result cache invalidation function was getting skipped in
case CPC is configured as SRAM.This was causing random crashes.
Signed-off-by: Ruchika Gupta <ruchika.gupta@freescale.com>
Signed-off-by: Aneesh Bansal <aneesh.bansal@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
Add NOR, SPI and SD secure boot targets for BSC9132QDS.
Changes:
- Debug TLB entry is not required for Secure Boot Target.
Signed-off-by: Aneesh Bansal <aneesh.bansal@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
For KVM we have a special PV machine type called "ppce500". This machine
is inspired by the MPC8544DS board, but implements a lot less features
than that one.
It also provides more PCI slots and is supposed to be enumerated by
device tree only.
This patch adds support for the generic ppce500 machine and tries to
rely solely on device tree for device enumeration.
Signed-off-by: Alexander Graf <agraf@suse.de>
Acked-by: Scott Wood <scottwood@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
There is no need to set IVORs to anything but their default values,
so let's leave them where they are.
Suggested-by: Scott Wood <scottwood@freescale.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
Acked-by: Scott Wood <scottwood@freescale.com>
[York Sun: Add back $(obj)start.S section in mpc8572ds/Makefile]
Reviewed-by: York Sun <yorksun@freescale.com>
Commit 643aae1406
deleted include/linux/config.h but missed to
delete _LINUX_CONFIG_H macro.
It is no longer used at all.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
In PBL RAMBOOT(SPI/SD/NAND boot) mode, CPC1 used as SRAM, should disable
CPC1 speculation and keep it till relocation. Otherwise, speculation
transactions will go to DDR controller, it will cause problem.
Signed-off-by: Dave Liu <daveliu@freescale.com>
Signed-off-by: Shaohui Xie <Shaohui.Xie@freescale.com>
Acked-by: York Sun <yorksun@freescale.com>
CHASSIS2 architecture never defines type of L2 cache present in SoC.
it is dependent upon the core present in the SoC.
for example,
- e6500 core has L2 cluster (Kibo)
- e5500 core has Backside L2 Cache
Signed-off-by: Prabhakar Kushwaha <prabhakar@freescale.com>
In a very rare condition, a system hang is possible when the e500 core
initiates a guarded load to PCI / PCIe /SRIO performs a coherent write
to memory. Please refer to errata document for more details. This erratum
applies to the following SoCs and their variants, if any.
BSC9132
BSC9131
MPC8536
MPC8544
MPC8548
MPC8569
MPC8572
P1010
P1020
P1021
P1022
P1023
P2020
C29x
Signed-off-by: York Sun <yorksun@freescale.com>
CC: Scott Wood <scottwood@freescale.com>
When a board (slave) boots from SRIO/PCIE, it would get the instructions
from a remote board (master) by SRIO/PCIE interface, and the slave's
u-boot image should be built with the
SYS_TEXT_BASE=0xFFF80000;
So the u-boot of the slave should avoid the NOR_BOOT branch at the
booting stage.
For example, when a P2041RDB boots from SRIO/PCIE, it will set TLB
entry 15 from base address "CONFIG_SYS_MONITOR_BASE & 0xffc00000",
and with the 4M size as the boot window in NOR_BOOT branch. Because
the CONFIG_SYS_MONITOR_BASE = CONFIG_SYS_TEXT_BASE = 0xFFF80000, so
the TLB entry will be from base address 0xffc00000 and with 4M size.
Then the u-boot will set TLB entry 14 from base address
"CONFIG_SYS_INIT_RAM_ADDR", and with the 16K size as the initial
stack window. For the P2041RDB platform, the CONFIG_SYS_INIT_RAM_ADDR
= 0xffd00000. So the TLB entry 14 and 15 will be in confliction.
There will be right TLB entries configurations when avoid the
NOR_BOOT branch and set the boot window from 0xfff00000 with 1M
size space.
Signed-off-by: Liu Gang <Gang.Liu@freescale.com>
There will clear the BSS in the function clear_bss(), the reset address of
the BSS started from the __bss_start, and increased by four-byte increments,
finally stoped depending on the address is equal to the _bss_end. If the end
address __bss_end is not alignment to 4byte, it will be an infinite loop.
1. The reset action stoped depending on the reset address is greater
than or equal the end address of the BSS.
2. The end address of the BSS should be 4byte aligned. Because the reset unit
is 4 Bytes.
This patch is on top of the patch "powerpc/mpc85xx: support application
without resetvec segment in the linker script".
Signed-off-by: Ying Zhang <b40530@freescale.com>
Signed-off-by: Andy Fleming <afleming@freescale.com>
dcbi instruction has been used to clear D-cache lock. However, the cache
lock is persistent for e6500 core. Use dcblc to clear the lock explicitly.
Signed-off-by: York Sun <yorksun@freescale.com>
Signed-off-by: Andy Fleming <afleming@freescale.com>
The L1 D-cache on e6500 is write-through. This means that it's not
considered a good idea to have the L1 up and running if the L2 is
disabled. We don't actually *use* the L1 until after the L2 is
brought up on e6500, so go ahead and move the L1 enablement after
that code is done.
Signed-off-by: Andy Fleming <afleming@freescale.com>
Makes it a bit easier to see if we've properly set them. While
we're in there, modify the accesses to HDBCR0 and HDBCR1 to actually
use those definitions.
Signed-off-by: Andy Fleming <afleming@freescale.com>
Note this is a tree-wide change affecting multiple architectures.
At present we use __bss_start, but mostly __bss_end__. This seems
inconsistent and in a number of places __bss_end is used instead.
Change to use __bss_end for the BSS end symbol throughout U-Boot. This
makes it possible to use the asm-generic/sections.h file on all
archs.
Signed-off-by: Simon Glass <sjg@chromium.org>
e6500 implements MMUv2 and supports power-of-2 page sizes rather than
power-of-4. Add support for such pages.
Signed-off-by: Scott Wood <scottwood@freescale.com>
Signed-off-by: Andy Fleming <afleming@freescale.com>
cpu_init_nand.c is renamed to spl_minimal.c as it is not really NAND-specific.
Signed-off-by: Scott Wood <scottwood@freescale.com>
---
v2: factor out START, and change cpu_init_nand.c to spl_minimal.c
Cc: Andy Fleming <afleming@freescale.com>
In the RAMBOOT/SPL case we were creating a TLB entry starting at
CONFIG_SYS_MONITOR_BASE, and just hoping that the base was properly
aligned for the TLB entry size. This turned out to not be the case
with NAND SPL because the main U-Boot starts at an offset into the image
in order to skip the SPL itself.
Fix the TLB entry to always start at a proper alignment. We still assume that
CONFIG_SYS_MONITOR_BASE doesn't start immediately before a large-page boundary
thus requiring multiple TLB entries.
Signed-off-by: Scott Wood <scottwood@frescale.com>
Cc: Andy Fleming <afleming@freescale.com>
Previously, in many if not all configs we were creating overlapping TLB entries
which is illegal. This caused a crash during boot when moving p2020rdb NAND SPL
into L2 SRAM.
Signed-off-by: Scott Wood <scottwood@freescale.com>
Cc: Prabhakar Kushwaha <prabhakar@freescale.com>
Cc: Andy Fleming <afleming@freescale.com>
--
Prabhakar, please test that debug still works.
Using E6500 L1 cache as initram requires L2 cache enabled.
Add l2-cache cluster enabling.
Setup stash id for L1 cache as (coreID) * 2 + 32 + 0
Setup stash id for L2 cache as (cluster) * 2 + 32 + 1
Stash id for L2 is only set for Chassis 2.
Signed-off-by: York Sun <yorksun@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Signed-off-by: Andy Fleming <afleming@freescale.com>
These assembly macros simplify codes to add and delete temporary TLB entries.
Signed-off-by: York Sun <yorksun@freescale.com>
Signed-off-by: Andy Fleming <afleming@freescale.com>
Instead of just shooting down the entry that covers CCSR, clear out
every TLB entry that isn't the one that we're executing out of.
Signed-off-by: Scott Wood <scottwood@freescale.com>
Signed-off-by: Andy Fleming <afleming@freescale.com>
