The hash command function were not flushing the dcache before passing data
to CAAM/DMA and not invalidating the dcache when getting data back.
Due the data cache incoherency, HW accelerated hash commands used to fail
with CAAM errors like "Invalid KEY Command".
Check if pbuf and pout buffers are properly aligned to the cache line size
and flush/invalidate the memory regions to address this issue.
This solution is based in a previous work from Clemens Gruber in
commit 598e9dccc7 ("crypto/fsl: fix BLOB encapsulation and
decapsulation")
Reported-by: Anatolij Gustschin <agust@denx.de>
Signed-off-by: Breno Lima <breno.lima@nxp.com>
Reviewed-by: York Sun <york.sun@nxp.com>
Sets DDR configuration parameter cdr1 before all other settings
to support case 0.9v VDD is enabled for some SoCs
Signed-off-by: Ashish Kumar <Ashish.Kumar@nxp.com>
Signed-off-by: Rajesh Bhagat <rajesh.bhagat@nxp.com>
Reviewed-by: York Sun <york.sun@nxp.com>
Judging from its name and parameters, device_is_compatible looks like it
is compatible with a live device tree, but it actually isn't.
Make it compatible with a live device tree.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Mario Six <mario.six@gdsys.cc>
There are some whitespace-related style violations in read.c; fix those.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Mario Six <mario.six@gdsys.cc>
The Makefile entries in the clk driver directory were not alphabetically
sorted. Correct this.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Mario Six <mario.six@gdsys.cc>
The clk uclass was converted to support a live device tree recently,
hence the global data pointer declarations are no longer needed.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Mario Six <mario.six@gdsys.cc>
checkpatch.pl complains that the clk_ops structures used in clk-uclass.c
ought to be const, so we mark them as const.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Mario Six <mario.six@gdsys.cc>
The function dev_read_addr calls ofnode_get_addr_index in the live tree
case, which does not apply bus translations to the address read from the
device tree. This results in illegal addresses on boards that rely on
bus translations being applied.
Fix this situation by applying bus translations in the live tree case as
well.
Signed-off-by: Mario Six <mario.six@gdsys.cc>
Tested-by: Stephen Warren <swarren@nvidia.com>
Now this driver is used across stm32f4, stm32f7 and stm32h7
SoCs family, give it a generic name.
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Reviewed-by: Vikas Manocha <vikas.manocha@st.com>
The command portion of the GPIO driver can only be used in full SPL so
re-work to guard the command related portions and mark it as static.
Cc: Dan Murphy <dmurphy@ti.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
The command portion of the GPIO driver can only be used in full SPL so
re-work to guard the command related portions and mark it as static.
Cc: Bin Meng <bmeng.cn@gmail.com>
Cc: Simon Glass <sjg@chromium.org>
Cc: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Add support for simple memory fill operation. With large data sizes
it is much faster to use EDMA for memory fill rather than CPU.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
omap_hsmmc driver uses "|" in a couple of places for disabling a bit.
While it's okay to use it in "mmc_reg_out" (since mmc_reg_out has a
_mask_ argument to take care of resetting a bit), it's incorrectly used
for resetting flags in "omap_hsmmc_send_cmd".
Fix it here by using "&= ~()" to reset a bit.
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Instead of sending STOP TRANSMISSION command from MMC core, enable
the auto command feature so that the Host Controller issues CMD12
automatically when last block transfer is completed.
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
The omap hsmmc host controller can have the ADMA2 feature. It brings better
read and write throughput.
On most SOC, the capability is read from the hl_hwinfo register. On OMAP3,
DMA support is compiled out.
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Update the calculation for tWR and tPD. This improves the DDR refresh
interval and brings the initialization into line with the binary blobs
currently being supplied by Marvell.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
Signed-off-by: Stefan Roese <sr@denx.de>
The RD_SAMPLE_DELAY field is 5 bits so it needs to be masked with 0x1f
instead of 0xf. Rather than checking the read sample delay for all DDR
chip selects use the values for the chip selects that are actually
configured. Finally continue searching for the max_phase value even if the
current read_sample is the same as the max_read_sample.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
Signed-off-by: Stefan Roese <sr@denx.de>
The ternary operation had the HIGH/LOW values the
wrong way round. Update it to use the correct value.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
Signed-off-by: Stefan Roese <sr@denx.de>
When using only a single DDR chip select only assert M_ODT[0] on write.
Do not assert it on read and do not assert M_ODT[1] at all. Also set
tODT_OFF_WR to 0x9 which contradicts the recommendation from the
functional spec but is what Marvell's binary training blob does and
seems to give better results when ODT is active during writes.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
Signed-off-by: Stefan Roese <sr@denx.de>
Update default AXI, TUN, ARC clock set options:
instead of changing only IDIV divider settings adjust also domain PLL
settings.
Add support of TUN_ROM and TUN_PWM clocks (subclocks of TUNN_PLL)
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
This reverts commit 46831c1a4c.
This reserved area at the beginning of struct hsmm, will be used later to
support ADMA
Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
The number 4 is dedicated on i.MX7ULP, but lpi2c will be reused on i.MX8,
4 is not valid. The seq number could be configured by alias node.
The following patch will use i2c4 as the begin for i.MX7ULP.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Identify and distinguish between platform device type of MX7ULP
and LS1021A.
This is a fix to commit 7edf5c45("serial: lpuart: add i.MX7ULP
support").
Signed-off-by: Sriram Dash <sriram.dash@nxp.com>
Acked-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: York Sun <york.sun@nxp.com>
This adds SH_ETHER to drivers/net/Kconfig and convert to Kconfig.
Signed-off-by: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
When we using network on board using sh-eth, it prints a lot of
"CACHE: Misaligned operation at range" messages.
This commit fixes this problem.
Signed-off-by: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
This changes Change structure used in sh_eth_read and sh_eth_write function
from struct sh_eth_dev to struct sh_eth_info. This is necessary to convert
to Driver Model.
Signed-off-by: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
bd_t is not used in sh_eth_config(). This deletes bd_t from sh_eth_config()
Signed-off-by: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
This fixes the chord style checked by checkpatch.pl.
Details of change details are as follows:
- Fix typo
Change from alligned to aligned.
- Remove whitespace before ','
- Add spaces preferred around that '|'
- Fix missing a blank line after declarations
- Remove space after a cast declaration
- Fix format of block comments
- Add a blank line after function/struct/union/enum declarations
Signed-off-by: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Add a b53_reg read/write command which allows inspecting the switch
registers. Because the Broadcom BCM53xx registers have different sizes,
we need to split the accesses in 8, 16, 32, 48 or 64 bits to obtain
expected results.
Reviewed-by: Stefan Roese <sr@denx.de>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Make sure that we pad small packets to a minimum length of 60 bytes
(without FCS). This is necessary to interface with Ethernet switches
that will reject RUNT frames unless padded correctly.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Add a minimalistic Broadcom BCM53xx (roboswitch) switch driver similar
to the Marvell MV88E617x. This takes care of configuring the minimum
amount out of the switch hardware such that each user visible port
(configurable) and the CPU port can forward packets between each other
while preserving isolation with other ports.
This is useful for e.g: the Lamobo R1 board featuring a Broadcom
BCM53125 switch.
Reviewed-by: Stefan Roese <sr@denx.de>
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
This implementation manages several clocks, disable and
free all of them in case of error during probe and in remove
callback.
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
This patch fixes a problem in the mvneta driver where if more than
one packet arrives between calls to mvneta_recv(), the additional
descriptors will be marked as free even though only one descriptor
has been read and processed from the receive queue. This causes
the additional packet(s) to be delayed until the next packet arrives.
>From this point on all packets will be delayed because the receive
queue will contain unprocessed packets but the hardware shows no
busy descriptors.
Signed-off-by: Jason Brown <jason.brown@apcon.com>
Reviewed-by: Stefan Roese <sr@denx.de>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Since the return value is a signed int, if the leading MSB of rdreg is a 1,
it will get signed extended and will return a negative value which is an
error even though we read the correct value.
Fixes: dfcc496ed7 ("net: mii: Changes not made by spatch")
Signed-off-by: Chris Brandt <chris.brandt@renesas.com>
Acked-by: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
First, this table could never be included in the build anyway because
SH_ETH_TYPE_RZ is not defined until later in the file.
Second, the register PIR was missing, so PHY MDIO never worked.
Third, after adding the PIR register, the table is EXACTLY the same as
sh_eth_offset_gigabit, so there is no value to it.
Therefore, just delete it use the gigabit one.
Signed-off-by: Chris Brandt <chris.brandt@renesas.com>
Acked-by: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
