Our early L1 relocate code may implicitly call code which lives in L1
memory. This is due to the dma_memcpy() rewrite that made the DMA code
lockless and safe to be used by multiple processes. If we start the
early DMA memcpy to relocate things into L1 instruction but then our
DMA memcpy code calls a function that lives in L1, things fall apart.
As such, create a small dedicated DMA memcpy routine that we can assume
sanity at boot time.
Reported-by: Filip Van Rillaer <filip.vanrillaer@oneaccess-net.com>
Signed-off-by: Robin Getz <robin.getz@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Random read/write errors are a bad thing - so don't let anyone
(including the test bench) run on something we know is bad.
Signed-off-by: Robin Getz <rgetz@blackfin.uclinux.org>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
rewrite get_sclk()/get_vco() based on the assumption sclk/vco never
changes (since today it cannot)
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
This is a mixture ofcMichael McTernan's patch and the existing cplb-mpu code.
We ditch the old cplb-nompu implementation, which is a good example of
why a good algorithm in a HLL is preferrable to a bad algorithm written in
assembly. Rather than try to construct a table of all posible CPLBs and
search it, we just create a (smaller) table of memory regions and
their attributes. Some of the data structures are now unified for both
the mpu and nompu cases. A lot of needless complexity in cplbinit.c is
removed.
Further optimizations:
* compile cplbmgr.c with a lot of -ffixed-reg options, and omit saving
these registers on the stack when entering a CPLB exception.
* lose cli/nop/nop/sti sequences for some workarounds - these don't
* make
sense in an exception context
Additional code unification should be possible after this.
[Mike Frysinger <vapier.adi@gmail.com>:
- convert CPP if statements to C if statements
- remove redundant statements
- use a do...while loop rather than a for loop to get slightly better
optimization and to avoid gcc "may be used uninitialized" warnings ...
we know that the [id]cplb_nr_bounds variables will never be 0, so this
is OK
- the no-mpu code was the last user of MAX_MEM_SIZE and with that rewritten,
we can punt it
- add some BUG_ON() checks to make sure we dont overflow the small
cplb_bounds array
- add i/d cplb entries for the bootrom because there is functions/data in
there we want to access
- we do not need a NULL trailing entry as any time we access the bounds
arrays, we use the nr_bounds variable
]
Signed-off-by: Michael McTernan <mmcternan@airvana.com>
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bernd Schmidt <bernds_cb1@t-online.de>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
If we are running on a chip revision below what we are compiled for,
there will be missing anomaly workarounds, and a panic is inevitable. Do
is sooner, rather than later, so people don't look for bugs that already
have workarounds (that they turned off).
Signed-off-by: Robin Getz <rgetz@blackfin.uclinux.org>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
do not allow people to pass in a diff clkin_hz value when
reprogramming clocks -- it is too late currently
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
rather than use *(unsigned int *)v everywhere, do this once with a local
cpu_num variable
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
- unify all dma in/out functions (takes ~35 lines of code now)
- unify dma_memcpy with dma in/out functions (1 place that touches MDMA0
registers)
- add support for 32bit transfers
- cleanup dma_memcpy code to be much more readable
- irqs are disabled only while programming MDMA registers rather than
the entire transaction
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
CCLK is variable: get current CCLK in show_cpuinfo
Signed-off-by: Michael Hennerich <michael.hennerich@analog.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
Blackfin dual core BF561 processor can support SMP like features.
https://docs.blackfin.uclinux.org/doku.php?id=linux-kernel:smp-like
In this patch, we provide SMP extend to Blackfin kernel and memory management code
Singed-off-by: Graf Yang <graf.yang@analog.com>
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
when copying L1 regions, go to the start of bss rather
than end since we have code to zero it out already
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
This patch introduces bfin_addr_dcachable() predicate, that simply tests is
address in cachable region or not.
Signed-off-by: Bryan Wu <cooloney@kernel.org>
Move all the silicon rev handling to one place (Kconfig) and
make sure we warn if you are running on silicon that has not been tested on
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
- remove cheesy read_iloc() function
- move invalidate_entire_icache function to lock.S
- export proper prototypes for functions in lock.S
- only build lock.S when BFIN_ICACHE_LOCK is enabled
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
use kernel command line mem and max_mem bootargs to limit
availabe memory instead.
Signed-off-by: Michael Hennerich <michael.hennerich@analog.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
detect the memory available in the system on the fly by default
rather than forcing people to set this manually in the kconfig
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
http://blackfin.uclinux.org/gf/project/uclinux-dist/tracker/?action=TrackerItemEdit&tracker_item_id=3978
Section data_l1_cacheline_aligned should be defined in
link script of kernel, when L1 data sram bank A is not available.
In bf536 with all data cache is enabled, there is no L1 data sram.
Current link script won't define section data_l1.cacheline_aligned in
this case. But, if user select put cacheline_aligned data into l1 sram
in kernel menuconfig, these data will be dropped and access to these
data will trigger data CPLB exception.
Do panic in l1 relocation code as well.
Signed-off-by: Sonic Zhang <sonic.zhang@analog.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
theres no need to declare ram{end,start,base} in the head.S files
when declaring them with the other memory related variables in setup.c
is so much simpler/nicer
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
Robin Getz