This commit adapts 4xx boards for partial linking with --gc-sections.
Signed-off-by: Stefan Roese <sr@denx.de>
Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Stefan Roese <sr@denx.de>
Cc: Tirumala Marri <tmarri@apm.com>
Cc: David Updegraff <dave@cray.com>
Cc: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
Cc: Dirk Eibach <eibach@gdsys.de>
Cc: Larry Johnson <lrj@acm.org>
Cc: Peter De Schrijver <p2@mind.be>
Cc: Niklaus Giger <niklaus.giger@netstal.com>
Cc: Daniel Poirot <dan.poirot@windriver.com>
Acked-by: Stefan Roese <sr@denx.de>
The switch from archive libraries to partial linking has introduced a
number of problems, that are non-trivial to solve. For example, it is
no longer possible to include individual object files in the linker
script as we did before for example in the case of boards with
embedded environment to fill up the gap caused by the need to align
the environment on flash erase block boundaries.
The best (but unfortunately not easiest) approach to address this
problem is to enable -ffunction-sections (and -fdata-sections) so
we can again (and even in much finer granularity) place certain code
where we want it. When doing this step, it seems only consequent to
also add --gc-sections which has the added benefit of reducing the
memory footprint of the U-Boot image (both in flash and in RAM).
Unfortunately, this requires changes to a lot of linker scripts.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Stefan Roese <sr@denx.de>
Cc: Kim Phillips <kim.phillips@freescale.com>
Cc: Andy Fleming <afleming@gmail.com>
Cc: Kumar Gala <galak@kernel.crashing.org>
Acked-by: Stefan Roese <sr@denx.de>
As we try to get rid of board specific config.mk files we must
provide a way for board specific settings of the LDSCRIPT variable
(path to the linker script) where needed.
We now implement the following hierarchy:
- Highest priority has a "#define CONFIG_SYS_LDCONFIG" in the board
config file.
- If CONFIG_SYS_LDCONFIG is not set, and the system is booting from
NAND (CONFIG_NAND_SPL is set), then a board specific linker
script board/$(BOARDDIR)/u-boot-nand.lds gets used.
- If we are not booting from NAND, we test if a processor specific
linker script arch/powerpc/cpu/$(CPU)/u-boot.lds exists; if so we
use that.
- As default, arch/powerpc/config.mk gets used.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Stefan Roese <sr@denx.de>
Cc: Kim Phillips <kim.phillips@freescale.com>
Cc: Kumar Gala <kumar.gala@freescale.com>
Cc: Andy Fleming <afleming@gmail.com>
Acked-by: Stefan Roese <sr@denx.de>
So far, only the BAB7xx board would call the initialise_w83c553f()
function for the WINBOND 83C553 chip, even though some other boards
(HIDDEN_DRAGON, Sandpoint8240, Sandpoint8245) enabled it in their
board configuration. These boards were also missing other config
settings needed for that, which resulted in build errors like this:
drivers/pci/libpci.o:(.got2+0x84): undefined reference to `ide_bus_offset'
Switch arch/powerpc/lib/board.c to call initialise_w83c553f() not on a
per-board base, but when a WINBOND_83C553 in enabled in a
configuration (like BAB7xx), and disable it in the boards that had
this set so far.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Yusdi Santoso <yusdi_santoso@adaptec.com>
Cc: Jim Thompson <jim@musenki.com>
Acked-by: Stefan Roese <sr@denx.de>
The start.S file was only half-rewritten for ELF relocations.
This bugfix completes the rewrite.
Signed-off-by: Albert Aribaud <albert.aribaud@free.fr>
bulk addition of ELF relocation support to ARM cpus
arm946es, arm720t,arm920t, arm925t, arm_intcm, ixp,
lh7a40x, s3c44b0, and sa1100.
Signed-off-by: Albert Aribaud <albert.aribaud@free.fr>
We explicitly link in the initcode.o in the Blackfin linker script, so
there is no need to merge it into the main common object for the linker
to pull in itself. This also fixes duplicate symbol errors with the new
partial linking logic.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
DECLARE_GLOBAL_DATA_PTR is currently defined within the scope
of function while it is a global pointer. Change the scope of
definition to replicate it's global scope. This seems to help
gcc 4.5 optimizations as well.
Signed-off-by: Nishanth Menon <nm@ti.com>
Signed-off-by: Sandeep Paulraj <s-paulraj@ti.com>
The Blackfin on-chip BootROM requires that fill operations (which is
used for the bss) be aligned to 4 bytes (base addr and total len).
Plus, the Blackfin early init asm code assumes the same thing. So
rather than making things work for no real gain, make sure the bss
len is padded to 4 bytes in the linker script.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
This lets the linker garbage collect these functions when they aren't
actually used by placing them into the standard .text.<func> section.
Signed-off-by: Mike Frysinger <vapier@gentoo.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>
older ld emitted all ELF relocations in input sections named
.rel.dyn, whereas newer ld uses names of the form .rel*. The
linker script only collected .rel.dyn input sections. Rewrite
to collect all .rel* input sections.
Signed-off-by: Albert Aribaud <albert.aribaud@free.fr>
The fixup procedure just stored a constant value in the
fixup table rather than just adjusting the table.
Although that doesn't seem to do any harm, it prevents
relocation more that once.
Signed-off-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se>
Use CONFIG_SYS_MONITOR_BASE instead of CONFIG_SYS_TEXT_BASE in early
init code so we can share the same code with NAND or NOR boot and not
have additional ifdefs in here.
Signed-off-by: Haiying Wang <Haiying.Wang@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Fix u-boot-nand.lds and u-boot-nand_spl.lds according to:
Author: Peter Tyser <ptyser@xes-inc.com>
Date: Wed Sep 29 14:05:56 2010 -0500
commit fbe53f59bd
85xx: Use gc-sections to reduce image size
Signed-off-by: Haiying Wang <Haiying.Wang@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
since commit 3667cbeed5
on beagle board the second sdram bank didn;t longer
work. Since this patch sdram settings just get copied
from bank a, but CMD_NOP, CMD_PRECHARGE, CMD_AUTOREFRESH
are not executed and after that mr register is also
not updated. This patch adds this for the bank b.
Signed-off-by: Heiko Schocher <hs@denx.de>
cc: Steve Sakoman <steve@sakoman.com>
cc: Sandeep Paulraj <s-paulraj@ti.com>
cc: Wolfgang Denk <wd@denx.de>
Signed-off-by: Sandeep Paulraj <s-paulraj@ti.com>
Commit c3d3a54 uses CONFIG_ARMV7 to determine whether to call the
v7_flush_cache_all function. This breaks the build for all non-OMAP3
boards (like Panda and OMAP4430SDP) since there is only a v7_flush_cache_all
implementation for OMAP3.
This patch uses CONFIG_OMAP3XXX instead of CONFIG_ARMV7 so that only boards
with a v7_flush_cache_all will make the call.
Tested on Beagle, Overo, Panda, and OMAP4430SDP
Signed-off-by: Steve Sakoman <steve.sakoman@linaro.org>
Signed-off-by: Sandeep Paulraj <s-paulraj@ti.com>
For ARM systems, before ELF relocation was introduced,
CONFIG_SKIP_RELOCATE_UBOOT coul be used to prevent *COPYING* the
U-Boot image from whereever it was loaded to it's link address
(CONFIG_SYS_TEXT_BASE). The name was badly chosen, as no relocation
was performed at all, it was just a memcpy().
With ELF relocation, this does not work like that any more, and
related boards need to be fixed anyway. So don't keep this relict any
longer.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Tested-by: Heiko Schocher <hs@denx.de>
Tested-by: Reinhard Meyer <u-boot@emk-elektronik.de>
When this define was introduced, the idea was to provide a soft
migration path for ARM boards to get adapted to the new relocation
support. However, other recent changes led to a different
implementation (ELF relocation), where this no longer works. By now
CONFIG_SYS_ARM_WITHOUT_RELOC does not only not help any more, but it
actually hurts because it obfuscates the actual code by sprinkling it
with lots of dead and non-working debris.
So let's make a clean cut and drop CONFIG_SYS_ARM_WITHOUT_RELOC.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Tested-by: Heiko Schocher <hs@denx.de>
Tested-by: Reinhard Meyer <u-boot@emk-elektronik.de>
By now, the majority of architectures have working relocation
support, so the few remaining architectures have become exceptions.
To make this more obvious, we make working relocation now the default
case, and flag the remaining cases with CONFIG_NEEDS_MANUAL_RELOC.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Tested-by: Heiko Schocher <hs@denx.de>
Tested-by: Reinhard Meyer <u-boot@emk-elektronik.de>
If start of any DRAM bank is greater than total DDR size, remaining DDR banks' start address & size were left un-initialized in dram_init function. This could break other functions who uses array 'gd->bd->bi_dram'. Kirkwood network driver is one example. This also stops Linux kernel from booting.
v2 - Set start address also to 0. Without this Linux kernel couldn't
boot up
Signed-off-by: Tanmay Upadhyay <tanmay.upadhyay@einfochips.com>
This patch is to fix build breakage and support new relocation
scheme for mx51evk.
- Correct IRAM base address and add size definition
The IRAM starts from 0x1FFE0000 on final revsion i.mx51 than
0x1FFE8000 which is for older revision.
- Include imx-regs.h in mx51evk.h
Definitions like CSD0_BASE_ADDR and IRAM_BASE_ADDR can be
referred to.
- Define CONFIG_SYS_INIT_RAM_ADDR and CONFIG_SYS_INIT_RAM_SIZE
They are used to define init RAM layout.
- Remove comment for CONFIG_SYS_GBL_DATA_SIZE which has been
buried by Wolfgang's commit below
25ddd1fb: Replace CONFIG_SYS_GBL_DATA_SIZE by auto-generated value
Signed-off-by: Shawn Guo <shawn.gsc@gmail.com>
The patch removes the warning:
clock.c:291: warning: initialization from incompatible pointer type
after constification of args[]
Signed-off-by: Stefano Babic <sbabic@denx.de>
This patch fixes the reset command on imx25. The watchdog registers are 16
bits in size and not 32. This patch also adds the service register codes as
constants.
Signed-off-by: Matthias Weisser <weisserm@arcor.de>
Remove u-boot.lds from mx5 and use the common u-boot.lds
of cpu layer. This patch also fix the building errors:
arch/arm/cpu/armv7/start.o: In function `_rel_dyn_start_ofs':
arch/arm/cpu/armv7/start.S:283: undefined reference to `__rel_dyn_start'
arch/arm/cpu/armv7/start.o: In function `_rel_dyn_end_ofs':
arch/arm/cpu/armv7/start.S:283: undefined reference to `__rel_dyn_end'
arch/arm/cpu/armv7/start.o: In function `_dynsym_start_ofs':
arch/arm/cpu/armv7/start.S:283: undefined reference to `__dynsym_start'
Signed-off-by: Jason Liu <r64343@freescale.com>
This patch adds generic support for the Samsung s3c2440 processor.
Global s3c24x0 changes to struct members converting from upper case to
lower case.
Signed-off-by: Craig Nauman <cnauman@diagraph.com>
Cc: kevin.morfitt@fearnside-systems.co.uk
Signed-off-by: Minkyu Kang <mk7.kang@samsung.com>
Now that the common code takes care of stripping away quotes and such
from numeric options, we no longer need to do so ourselves. So drop
the custom code we have in the Blackfin config.mk.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Since some of the defines in our config.h use the generated defines, we
need to include the generated header. This fixes building of the Blackfin
start.S file (where the stack is setup).
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
For the "fixloop" implementation in start.S a number of different
instructions was used. Unify code so all architectures use "blo"
here because it is more robust in case of incorrect alignments.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Albert ARIBAUD <albert.aribaud@free.fr>
Cc: Minkyu Kang <mk7.kang@samsung.com>
Cc: Sandeep Paulraj <s-paulraj@ti.com>
Cc: Prafulla Wadaskar <prafulla@marvell.com>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Marek Vasut <marek.vasut@gmail.com>
Acked-by: Heiko Schocher <hs@denx.de>
Fix relocation code for arm1176, do it like other ARM
CPU's are doing.
Tested only with CONFIG_SKIP_RELOCATE_UBOOT defined
and using nand_spl (booting from nand). Test done on
s3c6410 based board (not yet supported in main line).
Signed-off-by: Darius Augulis <augulis.darius@gmail.com>
Fix address setup bug for ARM.
This bug stops u-boot booting if
CONFIG_SKIP_RELOCATE_UBOOT is defined.
Signed-off-by: Darius Augulis <augulis.darius@gmail.com>
CONFIG_SYS_GBL_DATA_SIZE has always been just a bad workarond for not
being able to use "sizeof(struct global_data)" in assembler files.
Recent experience has shown that manual synchronization is not
reliable enough. This patch renames CONFIG_SYS_GBL_DATA_SIZE into
GENERATED_GBL_DATA_SIZE which gets automatically generated by the
asm-offsets tool. In the result, all definitions of this value can be
deleted from the board config files. We have to make sure that all
files that reference such data include the new <asm-offsets.h> file.
No other changes have been done yet, but it is obvious that similar
changes / simplifications can be done for other, related macro
definitions as well.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Acked-by: Kumar Gala <galak@kernel.crashing.org>
CONFIG_SYS_INIT_RAM_END was a misnomer as it suggests this might be
some end address; to make the meaning more clear we rename it into
CONFIG_SYS_INIT_RAM_SIZE
No other code changes are performed in this patch, only minor editing
of white space (due to the changed length) and the comments was done,
where noticed.
Note that the code for the PATI and cmi_mpc5xx board configurations
looks seriously broken. Last known maintainers on Cc:
Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Denis Peter <d.peter@mpl.ch>
Cc: Martin Winistoerfer <martinwinistoerfer@gmx.ch>
Acked-by: Kumar Gala <galak@kernel.crashing.org>
