Commit graph

18 commits

Author SHA1 Message Date
Sebastien Carlier
6d8962e814 Switch from archive libraries to partial linking
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>
2010-11-17 21:02:18 +01:00
Wolfgang Denk
14d0a02a16 Rename TEXT_BASE into CONFIG_SYS_TEXT_BASE
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>
2010-10-18 22:07:10 +02:00
Remy Bohmer
60f61e6d76 Convert DM9000 driver for CONFIG_NET_MULTI
All drivers need to be converted to CONFIG_NET_MULTI.
This patch converts the dm9000 driver.

Signed-off-by: Thomas Smits <ts.smits@gmail.com>
Signed-off-by: Remy Bohmer <linux@bohmer.net>
Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
2009-06-15 00:13:55 -07:00
Jean-Christophe PLAGNIOL-VILLARD
10a451cd57 arm: unify linker script
all arm boards except a few use the same cpu linker script
so move it to cpu/$(CPU)

that could be overwrite in following order
SOC
BOARD
via the corresponding config.mk

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
2009-06-12 20:39:52 +02:00
Trent Piepho
f62fb99941 Fix all linker script to handle all rodata sections
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>
2009-03-20 22:39:12 +01:00
Selvamuthukumar
9b827cf172 Align end of bss by 4 bytes
Most of the bss initialization loop increments 4 bytes
at a time. And the loop end is checked for an 'equal'
condition. Make the bss end address aligned by 4, so
that the loop will end as expected.

Signed-off-by: Selvamuthukumar <selva.muthukumar@e-coninfotech.com>
Signed-off-by: Wolfgang Denk <wd@denx.de>
2008-11-18 23:13:16 +01:00
Jean-Christophe PLAGNIOL-VILLARD
6d0f6bcf33 rename CFG_ macros to CONFIG_SYS
Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
2008-10-18 21:54:03 +02:00
Jean-Christophe PLAGNIOL-VILLARD
0e8d158664 rename CFG_ENV macros to CONFIG_ENV
Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
2008-09-10 22:48:06 +02:00
Wolfgang Denk
c8a3b109f0 Cleanup out-or-tree building for some boards (.depend)
Signed-off-by: Wolfgang Denk <wd@denx.de>
2008-07-02 23:49:18 +02:00
Jean-Christophe PLAGNIOL-VILLARD
db695b7851 scb9328: Fix flash warning: type qualifiers ignored on function return type
Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
2008-03-09 15:15:06 +01:00
Wolfgang Denk
64134f0112 Fix linker scripts: add NOLOAD atribute to .bss/.sbss sections
With recent toolchain versions, some boards would not build because
or errors like this one (here for ocotea board when building with
ELDK 4.2 beta):
ppc_4xx-ld: section .bootpg [fffff000 -> fffff23b] overlaps section .bss [fffee900 -> fffff8ab]

For many boards, the .bss section is big enough that it wraps around
at the end of the address space (0xFFFFFFFF), so the problem will not
be visible unless you use a 64 bit tool chain for development. On
some boards however, changes to the code size (due to different
optimizations) we bail out with section overlaps like above.

The fix is to add the NOLOAD attribute to the .bss and .sbss
sections, telling the linker that .bss does not consume any space in
the image.

Signed-off-by: Wolfgang Denk <wd@denx.de>
2008-01-12 20:31:39 +01:00
Wolfgang Denk
2b208f5308 Move "ar" flags to config.mk to allow for silent "make -s"
Based on patch by Mike Frysinger, 20 Jun 2006
2006-10-09 01:02:05 +02:00
Marian Balakowicz
f93286397e Add support for a saving build objects in a separate directory.
Modifications are based on the linux kernel approach and
support two use cases:

  1) Add O= to the make command line
  'make O=/tmp/build all'

  2) Set environement variable BUILD_DIR to point to the desired location
  'export BUILD_DIR=/tmp/build'
  'make'

The second approach can also be used with a MAKEALL script
'export BUILD_DIR=/tmp/build'
'./MAKEALL'

Command line 'O=' setting overrides BUILD_DIR environent variable.

When none of the above methods is used the local build is performed and
the object files are placed in the source directory.
2006-09-01 19:49:50 +02:00
Wolfgang Denk
d87080b721 GCC-4.x fixes: clean up global data pointer initialization for all boards. 2006-03-31 18:32:53 +02:00
Wolfgang Denk
807d5d7319 Fix problems with ld version 2.16 (dot outside sections problem)
Pointed out by Gerhard Jaeger, 31 Aug 2005;
cf. http://sourceware.org/ml/binutils/2005-08/msg00412.html
2005-08-31 12:28:00 +02:00
wdenk
400558b561 Prepare for SoC rework of ARM code:
- rename CONFIG_BOOTBINFUNC into  CONFIG_INIT_CRITICAL
- rename memsetup into lowlevel_init (function name and source files)
2005-04-02 23:52:25 +00:00
wdenk
731215ebde Patch by George G. Davis, 24 Aug 2004:
- update ARM boards to use constants from mach-types.h
2004-10-10 18:41:04 +00:00
wdenk
281e00a3be * Code cleanup
* Patch by Sascha Hauer, 28 Jun:
  - add generic support for Motorola i.MX architecture
  - add support for mx1ads, mx1fs2 and scb9328 boards

* Patches by Marc Leeman, 23 Jul 2004:
  - Add define for the PCI/Memory Buffer Configuration Register
  - corrected comments in cpu/mpc824x/cpu_init.c

* Add support for multiple serial interfaces
  (for example to allow modem dial-in / dial-out)
2004-08-01 22:48:16 +00:00