There may be a chance that board specific fix_fdt() will change the
size of FDT blob so it's safe to call reserve_fdt() after fix_fdt()
otherwise global data (gd) will overwrite with FDT blob values.
Fixes: a8492e25ac ("riscv: Expand the DT size before copy reserved memory node")
Signed-off-by: Pragnesh Patel <pragnesh.patel@sifive.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Rick Chen <rick@andestech.com>
Reviewed-by: Atish Patra <atish.patra@wdc.com>
Now that all arch specific code was converted to setup_bdinfo, we can
remove setup_board_part1.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Factor out ppc-specific bdinfo setup from generic init sequence to
arch_setup_bdinfo in arch/powerpc/lib/bdinfo.c.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Factor out m68k-specific bdinfo setup to arch_setup_bdinfo in
arch/m68k/lib/bdinfo.c. Also, use if(IS_ENABLED()) instead of #ifdef where
possible.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Move sram related bdinfo from arch-specific setup_board_part1 to generic
code in setup_bdinfo. Also use "if (IS_ENABLED(CONFIG_SYS_HAS_SRAM))"
instead of "#ifdef CONFIG_SYS_SRAM_BASE".
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Move all assignments to gd->bd->bi_mem{start,size} to generic code in
setup_bdinfo.
Xtensa architecture is special in this regard as it defines its own
handling of gd->bd->bi_mem{start,size} fields. In order to avoid defining
a weak SDRAM function, let arch_setup_bdinfo overwrite the generic flags.
For ARC architecture, remove ARCH_EARLY_INIT_R from Kconfig since it is
not needed anymore.
Also, use gd->ram_base to populate bi_memstart to avoid an ifdef.
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Acked-by: Alexey Brodkin <abrokdin@synopsys.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Certain architectures (ppc, mips, sh, m68k) use setup board_part1 and
setup_board_part2 calls during pre-relocation init to populate gd->bd
boardinfo fields. This makes the generic init sequence cluttered with
arch-specific ifdefs.
In order to clean these arch-specific sequences from generic init,
introduce arch_setup_bdinfo weak initcall so that everyone can define their
own bdinfo setup routines.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Introduce setup_bdinfo initcall as a generic routine to populate bdinfo
fields.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
The Linux coding style guide (Documentation/process/coding-style.rst)
clearly says:
It's a **mistake** to use typedef for structures and pointers.
Besides, using typedef for structures is annoying when you try to make
headers self-contained.
Let's say you have the following function declaration in a header:
void foo(bd_t *bd);
This is not self-contained since bd_t is not defined.
To tell the compiler what 'bd_t' is, you need to include <asm/u-boot.h>
#include <asm/u-boot.h>
void foo(bd_t *bd);
Then, the include direcective pulls in more bloat needlessly.
If you use 'struct bd_info' instead, it is enough to put a forward
declaration as follows:
struct bd_info;
void foo(struct bd_info *bd);
Right, typedef'ing bd_t is a mistake.
I used coccinelle to generate this commit.
The semantic patch that makes this change is as follows:
<smpl>
@@
typedef bd_t;
@@
-bd_t
+struct bd_info
</smpl>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
While relocating FDT we reserve some memory for the new FDT and
set the size of the FDT with that respect. But FDT may be placed
at the end of the RAM leading to memory access beyond it.
Fix this by copying exact FDT size bytes, not the reserved size.
Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
There is no technical reason to add additional 4k space for FDT. This space
is completely unused and just increase memory requirements. This is
problematic on systems with limited memory resources as Xilinx Zynq
CSE/ZynqMP mini and Versal mini configurations.
The patch is removing additional 4k space.
EFI code is using copy_fdt() which copy FDT to different location.
And all boot commands in case of using U-Boot's FDT pointed by
$fdtcontroladdr are copying FDT to different locations by
image_setup_libfdt().
That's why in proper flow none should modified DTB used by U-Boot that's
why there is no need for additional space.
Signed-off-by: Ashok Reddy Soma <ashok.reddy.soma@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Acked-by: Stephen Warren <swarren@nvidia.com>
Move this header out of the common header. Network support is used in
quite a few places but it still does not warrant blanket inclusion.
Note that this net.h header itself has quite a lot in it. It could be
split into the driver-mode support, functions, structures, checksumming,
etc.
Signed-off-by: Simon Glass <sjg@chromium.org>
CONFIG_SYS_SDRAM_BASE maybe zero. Avoid a build warning when compiling with
-Wtype-limits.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Current print_cpuinfo gets the first udevice in CPU class to return
the cpu info. This has problem if the boot CPU is not fixed.
Changing to use new API cpu_get_current_dev to fix the issue.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ye Li <ye.li@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Introduce arch_reserve_mmu to allow for architecture-specific reserve_mmu
routines. Also, define a weak nop stub for it.
Signed-off-by: Ovidiu Panait <ovpanait@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Move the ARM-specific reserve_mmu definition from common/board_f.c
to arch/arm/lib/cache.c.
Signed-off-by: Ovidiu Panait <ovpanait@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Add a function reserve_stack_aligned() to reserved memory with 16 bits
alignment after the stack pointer (gd->start_addr_sp) and use this new
function in board_f.c to reserve all the memory area (malloc, board, gd,
fdt, bootstage, stacks).
This 16 byte alignment is needed for cast on struct pointer
for the reserved memory, for example:
+ x86_64 ABI: https://reviews.llvm.org/D30049: 16 bytes
+ ARMv8 Instruction Set Overview: quad word, 16 bytes
An other alignment value could be needed for other architecture.
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Acked-by: Stephen Warren <swarren@nvidia.com>
In reserve_bootstage(), in case size is odd, gd->new_bootstage
is not aligned. In bootstage_relocate(), the platform hangs when
getting access to data->record[i].name.
To avoid this issue, make gd->new_bootstage 16 byte aligned.
To ensure that new_bootstage is 16 byte aligned (at least needed for
x86_64 and ARMv8) and new_bootstage starts down to get enough space,
ALIGN_DOWN macro is used.
Fixes: ac9cd4805c ("bootstage: Correct relocation algorithm")
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Reviewed-by: Vikas MANOCHA <vikas.manocha@st.com>
Reviewed-by: Patrick Delaunay <patrick.delaunay@st.com>
Tested-by: Patrick Delaunay <patrick.delaunay@st.com>
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
clear_bss is already used by 3 arches (x86, arc, xtensa), so make it generic
and provide a weak nop stub for it. This also removes arch-specific ifdef
duplications around clear_bss.
Signed-off-by: Ovidiu Panait <ovpanait@gmail.com>
This removes the arch-specific checks for "checkcpu" function from the init
sequence. Make "checkcpu" generic and provide a weak nop stub instead.
Signed-off-by: Ovidiu Panait <ovpanait@gmail.com>
At present panic() is in the vsprintf.h header file. That does not seem
like an obvious choice for hang(), even though it relates to panic(). So
let's put hang() in its own header.
Signed-off-by: Simon Glass <sjg@chromium.org>
[trini: Migrate a few more files]
Signed-off-by: Tom Rini <trini@konsulko.com>
These three clock functions don't use driver model and should be migrated.
In the meantime, create a new file to hold them.
Signed-off-by: Simon Glass <sjg@chromium.org>
Move this function into init.h which seems to be designed for this sort
of thing. Also update the header to declare struct global_data so that it
can be included without global_data.h being needed.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
At present bootstage relocation assumes that it is possible to point back
to memory available before relocation, so it does not relocate the
strings. However this is not the case on some platforms, such as x86 which
uses the cache as RAM and loses access to this when the cache is enabled.
Move the relocation step to before U-Boot relocates, expand the allocated
region to include space for the strings and relocate the strings at the
same time as the bootstage records.
This ensures that bootstage data can remain accessible from TPL through
SPL to U-Boot before/after relocation.
Signed-off-by: Simon Glass <sjg@chromium.org>
Make sure that the bloblist starts on an aligned boundary. This protects
against one of the early allocating causing the alignment to be lost.
Signed-off-by: Simon Glass <sjg@chromium.org>
The current code in reserve_noncached() has two issues:
1) The first update of gd->start_addr_sp always rounds down to a section
start. However, the equivalent calculation in cache.c:noncached_init()
always first rounds up to a section start, then subtracts a section size.
These two calculations differ if the initial value is already rounded to
section alignment.
2) The second update of gd->start_addr_sp subtracts exactly
CONFIG_SYS_NONCACHED_MEMORY, whereas the equivalent calculation in
cache.c:noncached_init() rounds the noncached size up to section
alignment before subtracting it. The two calculations differ if the
noncached region size is not a multiple of the MMU section size.
In practice, one/both of those issues causes a practical problem on
Jetson TX1; U-Boot triggers a synchronous abort during initialization,
likely due to overlapping use of some memory region.
This change fixes both these issues by duplicating the exact calculations
from noncached_init() into reserve_noncached().
However, this fix assumes that gd->start_addr_sp on entry to
reserve_noncached() exactly matches mem_malloc_start on entry to
noncached_init(). I haven't traced the code to see whether it absolutely
guarantees this in all (or indeed any!) cases. Consequently, I added some
comments in the hope that this condition will continue to be true.
Fixes: 5f7adb5b1c ("board_f: reserve noncached space below malloc area")
Cc: Vikas Manocha <vikas.manocha@st.com>
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Noncached area at present is being initialized to random space after malloc
area. It works in most the cases as it goes to stack area & stack is not
overwriting it being far from it.
Signed-off-by: Vikas Manocha <vikas.manocha@st.com>
This file contains lots of internal details about the environment. Most
code can include env.h instead, calling the functions there as needed.
Rename this file and add a comment at the top to indicate its internal
nature.
Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Reviewed-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
[trini: Fixup apalis-tk1.c]
Signed-off-by: Tom Rini <trini@konsulko.com>
This function relates to lcd.h and is about to become obsolete with the
driver-model conversion. Move it out of common.h
Signed-off-by: Simon Glass <sjg@chromium.org>
Correct the debug output for the trace buffer size to accommodate trace
buffers exceeding 2GiB.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Compiling with TRACE but without TRACE_EARLY results in an error
aarch64-linux-gnu-ld.bfd:
common/built-in.o:(.rodata.init_sequence_f+0x10):
undefined reference to `trace_early_init'
trace_early_init() should not be called if CONFIG_TRACE_EARLY is not
defined.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
While converting CONFIG_SYS_[DI]CACHE_OFF to Kconfig, there are instances
where these configuration items are conditional on SPL. This commit adds SPL
variants of these configuration items, uses CONFIG_IS_ENABLED(), and updates
the configurations as required.
Acked-by: Alexey Brodkin <abrodkin@synopsys.com>
Signed-off-by: Trevor Woerner <trevor@toganlabs.com>
[trini: Make the default depend on the setting for full U-Boot, update
more zynq hardware]
Signed-off-by: Tom Rini <trini@konsulko.com>
At present if one of the initcalls fails on sandbox the address printing
is not help, e.g.:
initcall sequence 0000557678967c80 failed at call 00005576709dfe1f (err=-96)
This is because U-Boot gets relocated high into memory and the relocation
offset (gd->reloc_off) does not work correctly for sandbox.
Add support for finding the base address of the text region (at least on
Linux) and use that to set the relocation offset. This makes the output
better:
initcall sequence 0000560775957c80 failed at call 0000000000048134 (err=-96)
Then you use can use grep to see which init call failed, e.g.:
$ grep 0000000000048134 u-boot.map
stdio_add_devices
Of course another option is to run it with a debugger such as gdb:
$ gdb u-boot
...
(gdb) br initcall.h:41
Breakpoint 1 at 0x4db9d: initcall.h:41. (2 locations)
Note that two locations are reported, since this function is used in both
board_init_f() and board_init_r().
(gdb) r
Starting program: /tmp/b/sandbox/u-boot
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
U-Boot 2018.09-00264-ge0c2ba9814-dirty (Sep 22 2018 - 12:21:46 -0600)
DRAM: 128 MiB
MMC:
Breakpoint 1, initcall_run_list (init_sequence=0x5555559619e0 <init_sequence_f>)
at /scratch/sglass/cosarm/src/third_party/u-boot/files/include/initcall.h:41
41 printf("initcall sequence %p failed at call %p (err=%d)\n",
(gdb) print *init_fnc_ptr
$1 = (const init_fnc_t) 0x55555559c114 <stdio_add_devices>
(gdb)
Signed-off-by: Simon Glass <sjg@chromium.org>
In legacy CONFIG_HARD_SPI initalizing spi_init code, which
was removed during dm conversion cleanup.
So remove the dead instances of CONFIG_HARD_SPI, and related
code.
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
spi_init used in some areas in tree, but the respective
drivers will remove in future patches.
So remove the same instances.
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
There is some basic informaton that SPL normally wants to pass through to
U-Boot, such as the SDRAM size and bank information.
Mkae use of the new bloblist structure for this. Add a new 'handoff' blob
which is set up in SPL and passed to U-Boot proper. Also adda test for
sandbox_spl that checks that this works correctly and a new 'sb' command
to show the information passed from SPL.
Reviewed-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Add support for locating a bloblist in U-Boot that has been set up by SPL.
It is copied into RAM during relocation.
Reviewed-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Convert the Watchdog driver for AT91SAM9x processors to support
the driver model and device tree. Changes "CONFIG_AT91SAM9_WATCHDOG"
to new "CONFIG_WDT_AT91" Kconfig option.
Signed-off-by: Prasanthi Chellakumar <prasanthi.chellakumar@microchip.com>
It was observed that current output of print_cpuinfo() on QEMU
x86 targets does not have an ending '\n', neither have a leading
'CPU:' any more. However it used to have these before.
It turns out commit c0434407b5 introduced a unified DM version
of print_cpuinfo() that exposed such issue on QEMU x86.
Fixes: c0434407b5 ("board_f: Use static print_cpuinfo if CONFIG_CPU is active")
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
When the DM CPU drivers are active, printing information about a CPU
should be delegated to a matching driver.
Hence, add a static print_cpuinfo that implements this delegation when
DM CPU drivers are active.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Mario Six <mario.six@gdsys.cc>
Changed condition to CONFIG_IS_ENABLED(CPU):
Signed-off-by: Simon Glass <sjg@chromium.org>
This reverts commit c0434407b5.
It turns out commit c0434407b5 broke some boards which have DM CPU
driver with CONFIG_DISPLAY_CPUINFO option on. These boards just fail
to boot when print_cpuinfo() is called during boot.
Fixes are already sent to ML and in u-boot-dm/next, however since
we are getting close to the v2018.11 release, it's safer we revert
the original commit.
This commit should be reverted after v2018.11 release.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
When the DM CPU drivers are active, printing information about a CPU
should be delegated to a matching driver.
Hence, add a static print_cpuinfo that implements this delegation when
DM CPU drivers are active.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Mario Six <mario.six@gdsys.cc>
Changed condition to CONFIG_IS_ENABLED(CPU):
Signed-off-by: Simon Glass <sjg@chromium.org>
To print the reset status during boot, add a method print_resetinfo to
board_f, which is called in init_sequence_f[], that gets the reset
information from the sysreset driver (assuming there is only one seems
reasonable), and prints it.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Mario Six <mario.six@gdsys.cc>
This patch updates the ram_base to store the start address of
the first bank DRAM and the use this ram_base to calculate ram_top
properly. This patch fixes the erroneous calculation of ram_top
incase of non zero ram start address.
Signed-off-by: Siva Durga Prasad Paladugu <siva.durga.paladugu@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
In case of no relocation we'll just waste some space at the very end
of usable memory area. If target device has very limited amount of memory
(for example 256 kB) this loss will be pretty inconvenient.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Cc: Bin Meng <bmeng.cn@gmail.com>
Cc: Heiko Schocher <hs@denx.de>
Cc: York Sun <york.sun@nxp.com>
Cc: Stefan Roese <sr@denx.de>
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>
Disabling relocation might be useful on ARC for 2 reasons:
a) For advanced debugging with Synopsys proprietary MetaWare debugger
which is capable of accessing much more specific hardware resources
compared to gdb. For example it may show contents of L1 and L2 caches,
internal states of some hardware blocks etc.
But on the downside MetaWare debugger still cannot work with PIE.
Even though that limitation could be work-arounded with change of ELF's
header and stripping down all debug info but with it we won't have
debug info for source-level debugging which is quite inconvenient.
b) Some platforms which might benefit from usage of U-Boot basically
don't have enough RAM to accommodate relocation of U-Boot so we
keep code in flash and use as much of RAM as possible for more
interesting things.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: Simon Glass <sjg@chromium.org>
Cc: Bin Meng <bmeng.cn@gmail.com>
Cc: Heiko Schocher <hs@denx.de>
Cc: York Sun <york.sun@nxp.com>
Cc: Stefan Roese <sr@denx.de>
Merge init_helpers.h in the new file init.h
with only prototypes for init_cache_f_r
used in common/board_f.c
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Move prototypes for function used in common/board_f.c
from common.h to init.h
Remove weak for arch_reserve_stacks in prototype
(checkpatch issue)
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Change all coldfire arch files to use CONFIG_DISPLAY_CPUINFO.
Signed-off-by: Angelo Dureghello <angelo@sysam.it>
Reviewed-by: Tom Rini <trini@konsulko.com>
---
Changes for v2:
- update common/Kconfig to add M68K to the default y list
Adds a VID specific API in init_sequence_f and spl code flow
namely init_func_vid which is required to adjust core voltage.
VID specific code is required in spl, hence moving flag CONFIG_VID
out of spl flags.
Signed-off-by: Ashish Kumar <Ashish.Kumar@nxp.com>
Signed-off-by: Rajesh Bhagat <rajesh.bhagat@nxp.com>
Reviewed-by: York Sun <york.sun@nxp.com>
Support common commands bdinfo and image format,
also modify common generic flow for RISC-V.
Signed-off-by: Rick Chen <rick@andestech.com>
Signed-off-by: Rick Chen <rickchen36@gmail.com>
Signed-off-by: Greentime Hu <green.hu@gmail.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
This does not appear to be used by any boards. Before introducing a new
log system, remove this old one.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Once U-Boot relocates itself the existing driver-model timer (if any) is
no-longer valid until the device is reinitialised. Any use of the device
may cause a crash. To handle this, set the timer to NULL after relocation.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
CONFIG_SYS_GENERIC_GLOBAL_DATA is no longer used by any board or
platform, so support for it can be dropped.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Vladimir Zapolskiy <vz@mleia.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
We are now using an env_ prefix for environment functions. Rename these
for consistency. Also add function comments in common.h.
Suggested-by: Wolfgang Denk <wd@denx.de>
Signed-off-by: Simon Glass <sjg@chromium.org>
Make reserve_mmu a weak so that it provides an option
to customize this routine as per platform need
Signed-off-by: Siva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
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>
Reviewed-by: Tom Rini <trini@konsulko.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>
U-Boot has up until now built with -fpic for the MIPS architecture,
producing position independent code which uses indirection through a
global offset table, making relocation fairly straightforward as it
simply involves patching up GOT entries.
Using -fpic does however have some downsides. The biggest of these is
that generated code is bloated in various ways. For example, function
calls are indirected through the GOT & the t9 register:
8f998064 lw t9,-32668(gp)
0320f809 jalr t9
Without -fpic the call is simply:
0f803f01 jal be00fc04 <puts>
This is more compact & faster (due to the lack of the load & the
dependency the jump has on its result). It is also easier to read &
debug because the disassembly shows what function is being called,
rather than just an offset from gp which would then have to be looked up
in the ELF to discover the target function.
Another disadvantage of -fpic is that each function begins with a
sequence to calculate the value of the gp register, for example:
3c1c0004 lui gp,0x4
279c3384 addiu gp,gp,13188
0399e021 addu gp,gp,t9
Without using -fpic this sequence no longer appears at the start of each
function, reducing code size considerably.
This patch switches U-Boot from building with -fpic to building with
-fno-pic, in order to gain the benefits described above. The cost of
this is an extra step during the build process to extract relocation
data from the ELF & write it into a new .rel section in a compact
format, plus the added complexity of dealing with multiple types of
relocation rather than the single type that applied to the GOT. The
benefit is smaller, cleaner, more debuggable code. The relocate_code()
function is reimplemented in C to handle the new relocation scheme,
which also makes it easier to read & debug.
Taking maltael_defconfig as an example the size of u-boot.bin built
using the Codescape MIPS 2016.05-06 toolchain (gcc 4.9.2, binutils
2.24.90) shrinks from 254KiB to 224KiB.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Cc: u-boot@lists.denx.de
Reviewed-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Tested-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Runtime U-boot dtb selection is generally a two step process. First step
is to simply use an initial generic dtb. The second step is to select
the dtb and perhaps execute additional code ones U-boot knows what board
it is running on. Embedded_dtb_select handles the second step by allowing
board specific code to run and perform what ever necessary configuration
that is needed.
Signed-off-by: Franklin S Cooper Jr <fcooper@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
There was for long time no activity in the 4xx area.
We need to go further and convert to Kconfig, but it
turned out, nobody is interested anymore in 4xx,
so remove it.
Signed-off-by: Heiko Schocher <hs@denx.de>
There was for long time no activity in the mpx5xxx area.
We need to go further and convert to Kconfig, but it
turned out, nobody is interested anymore in mpc5xxx,
so remove it.
Signed-off-by: Heiko Schocher <hs@denx.de>
relocate_code() calculates the relocation offset wrt. the symbol
__image_copy_start which happens to have the same value as
CONFIG_TEXT_BASE on most systems.
When creating an i.MX boot image with an integrated IVT it is
convenient to have CONFIG_TEXT_BASE point to the start of the IVT
that is prepended to the actual code. Thus CONFIG_TEXT_BASE will
differ from __image_copy_start, while the calculation
'gd->relocaddr - __image_copy_start' still gives the right relocation
offset.
Signed-off-by: Lothar Waßmann <LW@KARO-electronics.de>
The current implementation makes it look like the 'if (from_spl)' part is
dead code because these features are not enabled for sandbox. We could
enable it for sandbox_spl, but this is not done yet (it requires sharing
memory between SPL and U-Boot proper which is in fact supported).
It is probably nicer to avoid #ifdef anyway. Change it.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reported-by: Coverity (CID: 163244)
Fixes: 824bb1b (bootstage: Support SPL)
There was for long time no activity in the 5xx area.
We need to go further and convert to Kconfig, but it
turned out, nobody is interested anymore in 5xx,
so remove it.
Signed-off-by: Heiko Schocher <hs@denx.de>
There was for long time no activity in the 8260 area.
We need to go further and convert to Kconfig, but it
turned out, nobody is interested anymore in 8260,
so remove it.
Signed-off-by: Heiko Schocher <hs@denx.de>
There was for long time no activity in the 8xx area.
We need to go further and convert to Kconfig, but it
turned out, nobody is interested anymore in 8xx,
so remove it (with a heavy heart, knowing that I remove
here the root of U-Boot).
Signed-off-by: Heiko Schocher <hs@denx.de>
At present bootstage only supports U-Boot proper. But SPL can also consume
boot time so it is useful to have the record start there.
Add bootstage support to SPL. Also support stashing the timing information
when SPL finishes so that it can be picked up and reported by U-Boot
proper. This provides a full boot time record, excluding only the time
taken by the boot ROM.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present we don't allow use of bootstage before driver model is running.
This means we cannot time the init of driver model itself.
Now that bootstage requires its own board-specific timer, we can move its
init to earlier in the sequence, both before and after relocation.
Signed-off-by: Simon Glass <sjg@chromium.org>
Some boards cannot access pre-relocation data after relocation. Reserve
space for this and copy it during preparation for relocation.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present bootstage uses the data section of the image to store its
information. There are a few problems with this:
- It does not work on all boards (e.g. those which run from flash before
relocation)
- Allocated strings still point back to the pre-relocation data after
relocation
Now that U-Boot has a pre-relocation malloc() we can use this instead,
with a pointer to the data in global_data. Update bootstage to do this and
set up an init routine to allocate the memory.
Now that we have a real init function, we can drop the fake 'reset' record
and add a normal one instead.
Note that part of the problem with allocated strings remains. They are
reallocated but this will only work where pre-relocation memory is
accessible after relocation.
Signed-off-by: Simon Glass <sjg@chromium.org>
With a small tweak we can avoid including these files for all boards.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Coldfire arch is not happy with timer_init since interrupt handlers
are still not set at that stage, and the boot hangs silently.
Signed-off-by: Angelo Dureghello <angelo@sysam.it>
This allows us to use the same DRAM init function on all archs. Add a
dummy function for arc, which does not use DRAM init here.
Signed-off-by: Simon Glass <sjg@chromium.org>
[trini: Dummy function on nios2]
Signed-off-by: Tom Rini <trini@konsulko.com>
With d53ecad92f some unused interrupt related code was removed.
However all of these options are currently unused. Rather than migrate
some of these options to Kconfig we just remove the code in question.
The only related code changes here are that in some cases we use
CONFIG_STACKSIZE in non-IRQ related context. In these cases we rename
and move the value local to the code in question.
Fixes: d53ecad92f ("Merge branch 'master' of git://git.denx.de/u-boot-sunxi")
Signed-off-by: Tom Rini <trini@konsulko.com>
These includes don't seem to be needed now. Drop them. Reserve the
mp.h header for PowerPC for now.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Roese <sr@denx.de>
Reviewed-by: York Sun <york.sun@nxp.com>
This header file is used by three archs. It could be used by all of them
since relocation is a common function. Move it into a generic file.
Signed-off-by: Simon Glass <sjg@chromium.org>
This header file is used by two archs. It could be used by all of them
since it allows the cache to be on during relocation. Move it into a
generic file.
Signed-off-by: Simon Glass <sjg@chromium.org>
We don't need this PPC-specific function in generic code. Move it to
the powerpc directory.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Roese <sr@denx.de>
Drop headers which are not used or needed in this file. The compiler.h
header is included by common.h.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: York Sun <york.sun@nxp.com>
By making dram_init_banksize() return an error code we can drop the
wrapper. Adjust this and clean up all implementations.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Roese <sr@denx.de>
Move the ugly #ifdefs inside the reserve_video() function so we can
collect all this init into one place.
Signed-off-by: Simon Glass <sjg@chromium.org>
The arch-specific details of the cache being off are best handled inside
the reserve_mmu(). This cleans up the init sequence a little.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: York Sun <york.sun@nxp.com>
