Michal Simek [Fri, 26 Jan 2018 12:17:04 +0000 (13:17 +0100)]
fpga: Simplify error path in fpga_add
Check !desc earlier to simplify code.
Signed-off-by: Michal Simek <michal.simek@xilinx.com> Acked-by: Simon Goldschmidt <sgoldschmidt@de.pepperl-fuchs.com> Reviewed-by: Simon Goldschmidt <sgoldschmidt@de.pepperl-fuchs.com>
Lokesh Vutla [Fri, 16 Mar 2018 08:52:12 +0000 (14:22 +0530)]
davinci: Enable DDR_INIT for DA8XX
Commit 6aa4ad8e3820 ("Convert CONFIG_SOC_DA8XX et al to Kconfig")
converted SOC_DA8XX to Kconfig but missed enabling DDR_INIT for
SOC_DA8XX, which broke OMAPL138 to boot.
Commit 2e87980580d0 ("davinci: Fix omapl138_lcdk builds") disabled
DDR_INIT for all DA850 SoCs. This failed all DA850 boards to boot
as ddr is not being initialized.
Enable SYS_DA850_DDR_INIT for DA8XX so that all DA850 and OMAPL138
will have ddr initialized
Fixes: 2e87980580d0 ("davinci: Fix omapl138_lcdk builds") Fixes: 6aa4ad8e3820 ("Convert CONFIG_SOC_DA8XX et al to Kconfig") Reported-by: Sekhar Nori <nsekhar@ti.com> Tested-by: Sekhar Nori <nsekhar@ti.com> Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com> Reviewed-by: David Lechner <david@lechnology.com> Reviewed-by: Tom Rini <trini@konsulko.com>
Luca Ceresoli [Thu, 15 Mar 2018 10:08:56 +0000 (11:08 +0100)]
scripts/check-config.sh: fix "command not found" error handling
scripts/check-config.sh exits successfully and silently without doing
any checks when the 'comm' command is not found.
The problem triggers from the command around line 39:
comm -23 ${suspects} ${ok} >${new_adhoc}
This statement fails when 'comm' is not in $PATH, creating an empty
${new_adhoc} file. But the script continues and the following line,
which is supposed to detect an error:
if [ -s ${new_adhoc} ]; then
will always be false since the file is empty, and the script will exit
successfully as if everything were OK.
The case where 'comm' in not in $PATH is not theoretical. It used to
happen on yocto until a recent fix [0], and still happens on the
current stable branch (rocko).
Fix by setting the errexit flag to exit with error when a statement
fails, so that at least the problem is noticed.
For additional safety also set the nounset flag to detect expansion
errors.
David Lechner [Thu, 15 Mar 2018 01:36:30 +0000 (20:36 -0500)]
davinci: omapl138_lcdk: fix PLL0 frequency
commit 1601dd97edc6 ("davinci: omapl138_lcdk: increase PLL0 frequency")
changed the PLL0 frequency to 456MHz, which is needed for the LCDC IP
block. However, in doing so, it caused the PLLOUT clock to be outside
of the allowable specifications given in the OMAP-L138 data sheet. (It
says PLLOUT must be 600MHz max). It also uses a PLLM value outside of
the range given in the TRM (it says PLLM must in the range 0 to 0x1f).
Alexander Graf [Thu, 15 Mar 2018 14:10:20 +0000 (15:10 +0100)]
lan7xxx: Require phylib
The lan75xx and lan78xx drivers need to drive their phy via the generic
phylib framework. Let's reflect that dependency in Kconfig, so that we
don't get build errors when phylib does not get selected.
Signed-off-by: Alexander Graf <agraf@suse.de> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Alexander Graf [Thu, 15 Mar 2018 14:07:09 +0000 (15:07 +0100)]
net: Only access network devices after init
In the efi_loader main loop we call eth_rx() occasionally. This rx function
might end up calling into devices that haven't been initialized yet,
potentially resulting in a lot of transfer timeouts.
Instead, let's make sure the ethernet device is actually initialized before
reading from or writing to it.
Signed-off-by: Alexander Graf <agraf@suse.de> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
The "net_try_count" counter starts from "1".
And the "retrycnt" contains requested amount of retries.
With current logic, that means that the actual retry amount
will be one time less then what we set in "netretry" env.
For example setting "netretry" to "once" will make "retrycnt"
equal "1", so no retries will be triggered at all.
Fix the logic by changing the statement of "if" condition.
Signed-off-by: Leonid Iziumtsev <leonid.iziumtsev@se.atlascopco.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Priyanka Jain [Tue, 30 Jan 2018 06:41:08 +0000 (12:11 +0530)]
net/phy/cortina: Add No firmware upload option
Current Cortina phy driver assumes that firmware upload
is required during initialization and is dependent
on presence of corresponding macros like CONFIG_CORTINA_FW_ADDR
for compilation.
But Cortina phy has provision to store phy firmware in
attached dedicated EEPROM. And boards designed with such
EEPROM does not require firmware upload.
Add CORTINA_NO_FW_UPLOAD option in cortina.c to support
such boards.
Signed-off-by: Priyanka Jain <priyanka.jain@nxp.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Update get_phy_id() implementation in cortina.c to check
for Cortina_phy by comparing device phy_id with cortina phy_id
instead of relying on presence of CORTINA macros.
This will allow get_phy_id to work with non-cortina phy devices
which might have same phy address as Cortina device but on
different bus.
Signed-off-by: Priyanka Jain <priyanka.jain@nxp.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Calvin Johnson [Thu, 8 Mar 2018 10:00:33 +0000 (15:30 +0530)]
armv8: fsl-lsch2: configure pfe's DDR and HDBUS interfaces and ECC
1. Set AWCACHE0 attribute of PFE DDR and HDBUS master interfaces
to bufferable.
2. Set RD/WR QoS for PFE DDR and HDBUS AXI master interfaces.
3. Disable ECC detection for PFE.
Signed-off-by: Calvin Johnson <calvin.johnson@nxp.com> Signed-off-by: Anjaneyulu Jagarlmudi <anji.jagarlmudi@nxp.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Calvin Johnson [Thu, 8 Mar 2018 10:00:26 +0000 (15:30 +0530)]
drivers: net: pfe_eth: provide pfe commands
pfe_command provides command line support for several features that
support pfe, like starting or stopping the pfe, checking the health
of the processor engines and checking status of different units inside
pfe.
Signed-off-by: Calvin Johnson <calvin.johnson@nxp.com> Signed-off-by: Anjaneyulu Jagarlmudi <anji.jagarlmudi@nxp.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Following are the main driver files:-
pfe_hw.c: provides low level helper functions to initialize PFE
internal processor engines and other hardware blocks
pfe_driver.c: provides initialization functions
and packet send and receive functions
pfe_eth.c: provides high level gemac initialization functions
pfe_firmware.c: provides functions to load firmware into PFE
internal processor engines.
pfe_mdio.c: provides functions to initialize phy and mdio.
Signed-off-by: Calvin Johnson <calvin.johnson@nxp.com> Signed-off-by: Anjaneyulu Jagarlmudi <anji.jagarlmudi@nxp.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Calvin Johnson [Thu, 8 Mar 2018 10:00:23 +0000 (15:30 +0530)]
drivers: net: phy: Fix aquantia compilation with DM
With CONFIG_DM_ETH enabled, aquantia driver compilation fails with
below error. This patch fixes the issue by including dm.h.
drivers/net/phy/aquantia.c: In function ‘aquantia_startup’:
drivers/net/phy/aquantia.c:73:21: error: dereferencing pointer to
incomplete
type ‘struct udevice’
phydev->dev->name);
^~
Signed-off-by: Calvin Johnson <calvin.johnson@nxp.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
net: phy: Add PHY_RTL8211E_PINE64_GIGABIT_FIX for realtek phys
Setting PHY_RTL8211E_PINE64_GIGABIT_FIX forces internal rx/tx delays off
on the PHY, as well as flipping some magical undocumented bits. The
magic number comes from the Pine64 engineering team, presumably as a
proxy from Realtek. This configuration fixes the throughput on some
Pine64 models. Packet loss of up to 60-70% has been observed without
this.
Signed-off-by: Kyle Evans <kevans@FreeBSD.org> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Eugeniy Paltsev [Thu, 22 Mar 2018 10:50:47 +0000 (13:50 +0300)]
DW SPI: use 32 bit access instead of 16 and 32 bit mix
Current DW SPI driver uses 32 bit access for some registers and
16 bit access for others. So if DW SPI IP is connected via bus
which doesn't support 16 bit access we will get bus error.
Fix that by switching to 32 bit access only instead of 16 and 32 bit mix
Additional Documentation to Support this Change:
The DW_apb_ssi databook states:
"All registers in the DW_apb_ssi are addressed at 32-bit boundaries
to remain consistent with the AHB bus. Where the physical size of
any register is less than 32-bits wide, the upper unused bits of
the 32-bit boundary are reserved. Writing to these bits has no
effect; reading from these bits returns 0." [1]
[1] Section 6.1 of dw_apb_ssi.pdf (version 3.22a)
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com> Reviewed-by: Jagan Teki <jagan@openedev.com>
Eugeniy Paltsev [Thu, 22 Mar 2018 10:50:46 +0000 (13:50 +0300)]
DW SPI: add option to use external gpio for chip select
DW SPI internal chip select management has limitation:
it hold CS line in active state only when the FIFO is not
empty. If the FIFO freed before we add new data the SPI transaction will
be broken.
So add option to use external gpio for chip select. Gpio can be added
via device tree using standard gpio bindings.
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com> Reviewed-by: Jagan Teki <jagan@openedev.com>
Eugeniy Paltsev [Thu, 22 Mar 2018 10:50:45 +0000 (13:50 +0300)]
DW SPI: refactor poll_transfer functions
There is no sense in waiting for RX data in dw_reader function:
there is no chance that RX data will appear in RX FIFO if
RX FIFO is empty after previous TX write in dw_writer function.
So get rid of this waiting. After that we can get rid of dw_reader
return value and make it returning void. After that we can get rid
of dw_reader return value check in poll_transfer function.
With these changes we're getting closer to Linux DW SPI driver.
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com> Reviewed-by: Jagan Teki <jagan@openedev.com>
Eugeniy Paltsev [Thu, 22 Mar 2018 10:50:44 +0000 (13:50 +0300)]
DW SPI: fix transmit only mode
In current implementation we get -ETIMEDOUT error when we try to use
transmit only mode (SPI_TMOD_TO)
This happens because in transmit only mode input FIFO never gets any data
which breaks our logic in dw_reader(): we are waiting until RX data will be
ready in dw_reader, but this newer happens, so we return with error.
Fix that by using SPI_TMOD_TR instead of SPI_TMOD_TO which allows to use
RX FIFO.
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com> Reviewed-by: Jagan Teki <jagan@openedev.com>
Eugeniy Paltsev [Thu, 22 Mar 2018 10:50:43 +0000 (13:50 +0300)]
DW SPI: fix tx data loss on FIFO flush
In current implementation if some data still exists in Tx FIFO it
can be silently flushed, i.e. dropped on disabling of the controller,
which happens when writing 0 to DW_SPI_SSIENR (it happens in the
beginning of new transfer)
So add wait for current transmit operation to complete to be sure
that current transmit operation is finished before new one.
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com> Reviewed-by: Jagan Teki <jagan@openedev.com>
Tuomas Tynkkynen [Tue, 13 Mar 2018 13:23:04 +0000 (15:23 +0200)]
tools: Make kwboot build if HOST_TOOLS_ALL=y
The kwboot tool for Marvell devices isn't currently being built even if
HOST_TOOLS_ALL is set. It doesn't appear to depend on any CONFIG_
options, so it seems appropriate to enable building it here.
Signed-off-by: Tuomas Tynkkynen <tuomas@tuxera.com>
Eugeniy Paltsev [Wed, 21 Mar 2018 12:59:04 +0000 (15:59 +0300)]
ARC: Cache: Add missing cache cleanup before cache disable
Add missing cache cleanup before cache disable:
* Flush and invalidate L1 D$ before disabling. Flush and invalidate
SLC before L1 D$ disabling (as it will be bypassed for data)
Otherwise we can lose some data when we disable L1 D$ if this data
isn't flushed to next level cache. Or we can get wrong data if L1 D$
has some entries after enable which we modified when the L1 D$ was
disabled.
* Invalidate L1 I$ before disabling. Otherwise we can execute wrong
instructions after L1 I$ enable if we modified any code when
L1 I$ was disabled.
Eugeniy Paltsev [Wed, 21 Mar 2018 12:59:02 +0000 (15:59 +0300)]
ARC: Implement a function to sync and cleanup caches
Implement specialized function to clenup caches (and therefore
sync instruction and data caches) which can be used for cleanup before linux
launch or to sync caches during U-Boot self-relocation.
Eugeniy Paltsev [Wed, 21 Mar 2018 12:59:00 +0000 (15:59 +0300)]
ARC: Cache: Implement [i,d]cache_enabled() as separate functions
Implement icache_enabled() and dcache_enabled() as separate functions
which can be used with "inline" attribute. This is a preparation to
make them always_inline.
Eugeniy Paltsev [Wed, 21 Mar 2018 12:58:57 +0000 (15:58 +0300)]
ARC: Move cache global variables to arch_global_data
There is a problem with current implementation if we start U-Boot
from ROM, as we use global variables before ther initialization,
so these variables get overwritten when we copy .data section
from ROM.
Instead we move these global variables into our "global data"
structure so that we may really start from ROM.
Eugeniy Paltsev [Wed, 21 Mar 2018 12:58:56 +0000 (15:58 +0300)]
ARC: Cache: Get rid of [slc,pae,icache,dcache]_exists global variables
There is a problem with current implementation if we start U-Boot
from ROM, as we use global variables before ther initialization,
so these variables get overwritten when we copy .data section
from ROM.
Instead we'll use icache_exists(), dcache_exists(), slc_exists(), pae_exists()
functions which directly check BCRs every time.
In U-Boot case ops are used only during self-relocation and DMA
so we shouldn't be hit by noticeable performance degradation.
Eugeniy Paltsev [Wed, 21 Mar 2018 12:58:55 +0000 (15:58 +0300)]
ARC: Cache: Move SLC status check into slc_entire_op() and slc_rgn_op()
As of today we check SLC status before each call of __slc_rgn_op()
or __slc_entire_op(). So move status check into __slc_rgn_op()
and __slc_entire_op().
As we need to check status before *each* function execution and we
call slc_entire_op() and slc_rgn_op() from different places we add
this check directly into SLC entire/line functions instead of
their callers to avoid code duplication.
Eugeniy Paltsev [Wed, 21 Mar 2018 12:58:54 +0000 (15:58 +0300)]
ARC: Cache: Use is_isa_arcv2() instead of CONFIG_ISA_ARCV2 ifdef
Use is_isa_arcv2() function where it is possible instead of
CONFIG_ISA_ARCV2 define check to make code cleaner at the same time
keeping pretty much the same functionality - code in branches
under "if (is_isa_arcv2())" won't be compiled if CONFIG_ISA_ARCV2
is not defined, still we need a couple of CONFIG_ISA_ARCV2
ifdefs to make compiler happy. That's because code in
!is_isa_x() branch gets compiled and only then gets optimized
away.
Eugeniy Paltsev [Wed, 21 Mar 2018 12:58:53 +0000 (15:58 +0300)]
ARC: Cache: Allways check D$ status before entire/line ops
As we are planning to get rid of dozens of ifdef's in cache.c we
would better check D$ status before each entire/line operation
then check CONFIG_SYS_DCACHE_OFF config option.
This makes the code cleaner as well as D$ entire/line functions
remain functional even if we enable or disable D$ in run-time.
As we need to check status before *each* function execution and we
call D$ entire/line functions from different places we add
this check directly into D$ entire/line functions instead of
their callers to avoid code duplication.
Eugeniy Paltsev [Wed, 21 Mar 2018 12:58:52 +0000 (15:58 +0300)]
ARC: Move BCR encodings to separate header file
We're starting to use more and more BCRs and having their
definitions in-lined in sources becomes a bit annoying
so we move it all to a separate header.
Eugeniy Paltsev [Wed, 21 Mar 2018 12:58:50 +0000 (15:58 +0300)]
ARC: Flush & invalidate D$ with a single command
We don't implement separate flush_dcache_all() intentionally as
entire data cache invalidation is dangerous operation even if we flush
data cache right before invalidation.
There is the real example:
We may get stuck in the following code if we store any context (like
BLINK register) on stack in invalidate_dcache_all() function.
BLINK register is the register where return address is automatically saved
when we do function call with instructions like 'bl'.
void flush_dcache_all() {
__dc_entire_op(OP_FLUSH);
// Other code //
}
void invalidate_dcache_all() {
__dc_entire_op(OP_INV);
// Other code //
}
Now let's see what really happens during that code execution:
foo()
|->> call flush_dcache_all
[return address is saved to BLINK register]
[push BLINK] (save to stack) ![point 1]
|->> call __dc_entire_op(OP_FLUSH)
[return address is saved to BLINK register]
[flush L1 D$]
return [jump to BLINK]
<<------
[other flush_dcache_all code]
[pop BLINK] (get from stack)
return [jump to BLINK]
<<------
|->> call invalidate_dcache_all
[return address is saved to BLINK register]
[push BLINK] (save to stack) ![point 2]
|->> call __dc_entire_op(OP_FLUSH)
[return address is saved to BLINK register]
[invalidate L1 D$] ![point 3]
// Oops!!!
// We lose return address from invalidate_dcache_all function:
// we save it to stack and invalidate L1 D$ after that!
return [jump to BLINK]
<<------
[other invalidate_dcache_all code]
[pop BLINK] (get from stack)
// we don't have this data in L1 dcache as we invalidated it in [point 3]
// so we get it from next memory level (for example DDR memory)
// but in the memory we have value which we save in [point 1], which
// is return address from flush_dcache_all function (instead of
// address from current invalidate_dcache_all function which we
// saved in [point 2] !)
return [jump to BLINK]
<<------
// As BLINK points to invalidate_dcache_all, we call it again and
// loop forever.
Fortunately we may do flush and invalidation of D$ with a single one
instruction which automatically mitigates a situation described above.
And because invalidate_dcache_all() isn't used in common U-Boot code we
implement "flush and invalidate dcache all" instead.
Eugeniy Paltsev [Wed, 21 Mar 2018 12:58:49 +0000 (15:58 +0300)]
ARC: Introduce is_isa_X() functions
Introduce is_isa_arcv2() and is_isa_arcompact() functions.
These functions only check configuration options and return
compile-time constant so they can be used instead of #ifdef's to
to write cleaner code.
Now we can write:
-------------->8---------------
if (is_isa_arcv2())
ioc_configure();
-------------->8---------------
instead of:
-------------->8---------------
ifdef CONFIG_ISA_ARCV2
ioc_configure();
endif
-------------->8---------------
Eugeniy Paltsev [Wed, 21 Mar 2018 12:58:47 +0000 (15:58 +0300)]
ARC: Cache: Remove per-line I$ operations as unused
__cache_line_loop() function was copied from Linux kernel
where per-line instruction cache operations are really used.
In U-Boot we use only entire I$ ops, so we can drop support of
per-line I$ ops from __cache_line_loop() because __cache_line_loop()
is never called with OP_INV_IC parameter.
Alexey Brodkin [Wed, 21 Feb 2018 09:58:00 +0000 (12:58 +0300)]
arc: Fine-tune implementation of memory barriers
We improve on 2 things:
1. Only ARC HS family has "dmb" instructions so do compile-time
check for automatically defined macro __ARCHS__.
Previous check for ARCv2 ISA was not good enough because ARC EM
family is v2 ISA as well but still "dmb" instaruction is not
supported in EM family.
2. Still if there's no dedicated instruction for memory barrier
let's at least insert compile-time barrier to make sure
compiler deosn't reorder critical memory operations.
Alexey Brodkin [Wed, 16 Dec 2015 16:24:10 +0000 (19:24 +0300)]
arc: Introduce a possibility to not relocate U-boot
Disabling relocation might be useful on ARC for 2 reasons:
a) For advanced debugging with Synopsys proprietary MetaWare debugger
which is capable of accessing much more specific hardware resources
compared to gdb. For example it may show contents of L1 and L2 caches,
internal states of some hardware blocks etc.
But on the downside MetaWare debugger still cannot work with PIE.
Even though that limitation could be work-arounded with change of ELF's
header and stripping down all debug info but with it we won't have
debug info for source-level debugging which is quite inconvenient.
b) Some platforms which might benefit from usage of U-Boot basically
don't have enough RAM to accommodate relocation of U-Boot so we
keep code in flash and use as much of RAM as possible for more
interesting things.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com> Cc: Simon Glass <sjg@chromium.org> Cc: Bin Meng <bmeng.cn@gmail.com> Cc: Heiko Schocher <hs@denx.de> Cc: York Sun <york.sun@nxp.com> Cc: Stefan Roese <sr@denx.de>
Alexey Brodkin [Wed, 24 Jan 2018 12:27:04 +0000 (15:27 +0300)]
arc: Don't halt slaves
This commit basically reverts two commits:
1. cf628f772ef2 ("arc: arcv1: Disable master/slave check")
2. 6cba327bd96f ("arcv2: Halt non-master cores")
With mentioned commits in-place we experience more trouble than
benefits. In case of SMP Linux kernel this is really required as
we have all the cores running from the very beginning and then we
need to allow master core to do some preparatory work while slaves
are not getting in the way.
In case of U-Boot we:
a) Don't really run more than 1 core in parallel
b) We may use whatever core for that
projects / u-boot / log