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>
We cannot rely on the syscontrol rom func to program PLL_DIV with anomaly
05000440 is in effect, so manually program the MMR when necessary.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The intention all along was to accept pin names irrelevant of their case.
But I guess I forgot to test/implement support for that.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Make the GPIO command usable in a scripting environment by returning
the GPIO value rather than always 0.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The input sub command was missing from the help text, and it didn't show
the actual value currently read on the GPIO. This allows people to read
the value of input pins.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Serial devices currently have to manually stuff \r after every \n found,
but this is a bit more difficult with the jtag console since we process
everything in chunks of 4 bit. So we have to scan & stuff the whole
string rather than what most serial drivers do which is output on a byte
per byte basis.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
While we're in here, add some useful debug points. We need custom debug
statements because we need the output to only go to the serial port. If
we used the standard debug helpers, the output would also go to the stdout
(which would be the jtag console) and make it hard to figure out what is
going where exactly.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
If the other side isn't listening, we should reset the state to ignore
the whole message and not just the part we missed. This makes it easier
to connect at any time to the jtag console without worrying about the two
sides getting out of sync and thus sending garbage back and forth.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Use the new portmux framework to handle the details when possible.
Unfortunately, we cannot yet use this in the standalone initialization
logic, so we need to keep around the old portmux writes for now.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Now that we have a unified gpio layer, the misc partial gpio commands
can be unified and made complete (support all possible gpios).
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The current pinmux handling has spread throughout Blackfin drivers and
board code and is getting hideous to maintain. So import the gpio and
portmux layer from the Blackfin Linux code. This should spur a serious
of cleanups across the Blackfin tree.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Older on-chip Blackfin bootroms do not create a dummy NMI handler, so set
up one ourselves when anomaly 05000219 applies.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The default storage location for bootcount is EVT0. This version uses
one 32bit value and combines the magic/count value in the upper/lower
16bits. If there is demand for more, should be easy to do.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Avoid banging on the trace MMRs when debugging is disabled, avoid calling
the funcs multiple times in a row, disable the trace buffer earlier in the
exception handler to avoid eating more user entries, and dump the buffer
before calling the kgdb hook. This way we maximize useful debugging info
up front rather than needing external tools (like gdb/serial/etc...).
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The hush shell dynamically allocates (and re-allocates) memory for the
argument strings in the "char *argv[]" argument vector passed to
commands. Any code that modifies these pointers will cause serious
corruption of the malloc data structures and crash U-Boot, so make
sure the compiler can check that no such modifications are being done
by changing the code into "char * const argv[]".
This modification is the result of debugging a strange crash caused
after adding a new command, which used the following argument
processing code which has been working perfectly fine in all Unix
systems since version 6 - but not so in U-Boot:
int main (int argc, char **argv)
{
while (--argc > 0 && **++argv == '-') {
/* ====> */ while (*++*argv) {
switch (**argv) {
case 'd':
debug++;
break;
...
default:
usage ();
}
}
}
...
}
The line marked "====>" will corrupt the malloc data structures and
usually cause U-Boot to crash when the next command gets executed by
the shell. With the modification, the compiler will prevent this with
an
error: increment of read-only location '*argv'
N.B.: The code above can be trivially rewritten like this:
while (--argc > 0 && **++argv == '-') {
char *arg = *argv;
while (*++arg) {
switch (*arg) {
...
Signed-off-by: Wolfgang Denk <wd@denx.de>
Acked-by: Mike Frysinger <vapier@gentoo.org>
