When gcc compiles some 64 bit operations on a 32 bit machine, it generates
calls to small functions instead of instructions which do the job directly.
Those functions are defined in libgcc and transparently provide whatever
functionality was necessary. Unfortunately, u-boot can be built with a
non-standard ABI when libgcc isn't. More specifically, u-boot uses
-mregparm. When the u-boot and libgcc are linked together, very confusing
bugs can crop up, for instance seemingly normal integer division or modulus
getting the wrong answer or even raising a spurious divide by zero
exception.
This change borrows (steals) a technique and some code from coreboot which
solves this problem by creating wrappers which translate the calling
convention when calling the functions in libgcc. Unfortunately that means
that these instructions which had already been turned into functions have
even more overhead, but more importantly it makes them work properly.
To find all of the functions that needed wrapping, u-boot was compiled
without linking in libgcc. All the symbols the linker complained were
undefined were presumed to be the symbols that are needed from libgcc.
These were a subset of the symbols covered by the coreboot code, so it was
used unmodified.
To prevent symbols which are provided by libgcc but not currently wrapped
(or even known about) from being silently linked against by code generated
by libgcc, a new copy of libgcc is created where all the symbols are
prefixed with __normal_. Without being purposefully wrapped, these symbols
will cause linker errors instead of silently introducing very subtle,
confusing bugs.
Another approach would be to whitelist symbols from libgcc and strip out
all the others. The problem with this approach is that it requires the
white listed symbols to be specified three times, once for objcopy, once so
the linker inserts the wrapped, and once to generate the wrapper itself,
while this implementation needs it to be listed only twice. There isn't
much tangible difference in what each approach produces, so this one was
preferred.
Signed-off-by: Gabe Black <gabeblack@chromium.org>
U-Boot Makefiles contain a number of tests for compiler features etc.
which so far are executed again and again. On some architectures
(especially ARM) this results in a large number of calls to gcc.
This patch makes sure to run such tests only once, thus largely
reducing the number of "execve" system calls.
Example: number of "execve" system calls for building the "P2020DS"
(Power Architecture) and "qong" (ARM) boards, measured as:
-> strace -f -e trace=execve -o /tmp/foo ./MAKEALL <board>
-> grep execve /tmp/foo | wc -l
Before: After: Reduction:
==================================
P2020DS 20555 15205 -26%
qong 31692 14490 -54%
As a result, built times are significantly reduced, typically by
30...50%.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Andy Fleming <afleming@gmail.com>
Cc: Kumar Gala <galak@kernel.crashing.org>
Cc: Albert Aribaud <albert.aribaud@free.fr>
cc: Graeme Russ <graeme.russ@gmail.com>
cc: Mike Frysinger <vapier@gentoo.org>
Tested-by: Graeme Russ <graeme.russ@gmail.com>
Tested-by: Matthias Weisser <weisserm@arcor.de>
Tested-by: Sanjeev Premi <premi@ti.com>
Tested-by: Simon Glass <sjg@chromium.org>
Tested-by: Macpaul Lin <macpaul@gmail.com>
Acked-by: Mike Frysinger <vapier@gentoo.org>
Currently, some linker scripts are found by common code in config.mk.
Some are found using CONFIG_SYS_LDSCRIPT, but the code for that is
sometimes in arch config.mk and sometimes in board config.mk. Some
are found using an arch-specific rule for looking in CPUDIR, etc.
Further, the powerpc config.mk rule relied on CONFIG_NAND_SPL
when it really wanted CONFIG_NAND_U_BOOT -- which covered up the fact
that not all NAND_U_BOOT builds actually wanted CPUDIR/u-boot-nand.lds.
Replace all of this -- except for a handful of boards that are actually
selecting a linker script in a unique way -- with centralized ldscript
finding.
If board code specifies LDSCRIPT, that will be used.
Otherwise, if CONFIG_SYS_LDSCRIPT is specified, that will be used.
If neither of these are specified, then the central config.mk will
check for the existence of the following, in order:
$(TOPDIR)/board/$(BOARDDIR)/u-boot-nand.lds (only if CONFIG_NAND_U_BOOT)
$(TOPDIR)/$(CPUDIR)/u-boot-nand.lds (only if CONFIG_NAND_U_BOOT)
$(TOPDIR)/board/$(BOARDDIR)/u-boot.lds
$(TOPDIR)/$(CPUDIR)/u-boot.lds
Some boards (sc3, cm5200, munices) provided their own u-boot.lds that
were dead code, because they were overridden by a CPUDIR u-boot.lds under
the old powerpc rules. These boards' own u-boot.lds have bitrotted and
no longer work -- these lds files have been removed.
Signed-off-by: Scott Wood <scottwood@freescale.com>
Tested-by: Graeme Russ <graeme.russ@gmail.com>
