Add a set of helper functions for the CDC library that creates the correct
constraints for the CDC blocks. This makes it easier to specify the
constraints in the individual user's SDC files.
This only works for Altera where full scripting capabilities are available
in the SDC files.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Allow to specify additional properties when defining a IP parameter. The
properties take the form of a list of key value pairs. E.g.
ad_ip_parameter ... { \
DISPLAY_NAME "Name" \
DISPLAY_HINT "radio" \
}
This helps to reduce the amount of boilerplate when additional properties
need to be specified for a parameter.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
TCL files can be helpful to automate certain tasks like creating timing
constraints. Add handling for them to the ad_ip_addfile function.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Currently the reset for the link clock domain is generated internally in
the axi_jesd204_{rx,tx} peripheral. The reset is controlled by through the
register map.
Add an additional external reset for link clock domain. The link clock
domain is kept in reset if either the internal reset or the external reset
is asserted.
This for example allows the fabric to keep the domain in reset if the clock
is not yet stable.
The status of the external reset can be queried from the register map.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
dma_raddr is only incremented if it is less than dma_waddr_rel_s.
dma_waddr_rel_s is always less or equal to adc_waddr_rel << RATIO and
adc_waddr_rel is less than DMA_ADDR_LIMIT >> RATIO.
By induction we can conclude that this means that dma_raddr will always be
less then DMA_ADDR_LIMIT and the check for this will always evaluate to
false can be removed.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
When the DMA clock to ADC data rate ratio exceeds a certain threshold it is
possible that an erroneous dma_waddr_rel toggle event is generated. This
causes the last address of the previous DMA transfer to be transferred to
the DMA domain. And the DMA side will start reading from the FIFO even
though data is not available yet.
This results in data corruption with the current transfer containing data
from the previous transfer.
The root cause here is that the toggle signal CDC synchronizer register are
reset in the DMA when a new transfer starts, but not in the ADC domain,
causing a potential mismatch and the incorrect toggle event. To fix this
remove the reset from the DMA side. This is OK since the registers are
self-resetting if the reset signal is asserted for more than 3 clock cycles.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
AND logic means that all enabled triggers need to evaluate to true, others
are don't care. Fix the logic to behave accordingly.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
At the moment the drain signal is always asserted when the controller is
enabled. This breaks backpressure and data is lost. The drain signal should
only be asserted when the controller gets disabled until the last beat of
the current DMA transfer.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The 'PRIMITIVE_SUBGROUP == flop' filter only works on 7-Series. Replace it
with 'IS_SEQUENTIAL' which works on both 7-Series and UltraScale.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
In its default configuration the ram_2port module as a read latency of 2
clock cycles. Both the read address as well as the output data are
registered.
This is not the behavior that is expected from the alt_mem_asym module and
causes incorrect behavior and data corruption in the util_adc_fifo.
Disable the data output register to get a read latency of 1 clock cycle.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Name all CDC blocks following the patter i_cdc_${signal_name}. This makes
it clear what is going on.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Use the CDC sync_bits helper to synchronize the asynchronous external SYNC~
signal into the link clock domain, rather than open-coding this operation.
This makes it more explicit what is going on.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Which events will be exposed as IRQs and at what level of granularity will
need some additional through. Remove the two existing IRQ events for now
again. This will be added back later.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The up_cfg_ilas_data signal is a two dimensional array. There are 4
register entries for each lane. Model it as such rather than compressing it
down to a one dimensional array. This makes accessing the individual
entries a bit more straight forward and the code clearer.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The ilas_cfg_static.v is part of the jesd204_tx_static_config module.
Somehow a copy of that file made it into the jesd204_tx module where it is
completely unused. Remove the duplicated file.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
We know that NUM_OF_LANES will never exceed 255, but the tools don't know
and generate a warning about implicit signal truncation. Make the
truncation explicit to indicate that this is intentional.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
+ update to verilog-2001 coding standard
+ define RATIO outside the generate block
+ $clog2 macro is not supported by some tools, define function
locally
The CIC filter introduces different amplifications depending on the
decimation ratio. By adding a multiplier in the decimation chain
the amplification can be compensated
The ADI transport layer peripherals expect the first octet to be in the
LSBs and the last octet to be in the MSBs. The Altera JESD204 core orders
the octets the other way around though, first octet in the MSBs and last
octet in the LSBS.
Currently this is handled by having each transport layer peripheral swap
the octets around when it is connected to the Altera JESD204 core.
Change this so that rather than having to do the data swizzling in every in
every transport layer peripheral perform it at the input/output of the link
layer peripheral inside the generated block.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Currently the TX lane mapping is implemented by having to connect tx_phy_s_* to
the tx_ip_s_* and the tx_phy_d_* to the tx_ip_d_* signals in the system
qsys file in the desired order.
Re-work things so that instead the lane mapping is provided through the
TX_LANE_MAP parameter. The parameter specifies in which order logical lanes
are mapped onto the physical lanes.
The appropriate connections are than made inside the core according to this
parameter rather than having to manually connect the signals externally.
In order to generate a 1-to-1 mapping the TX_LANE_MAP parameter can be left
empty.
This change slightly reduces the boiler-plate code that is necessary to
setup the transceiver.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The CIC filter introduces different amplifications depending on the
interpolation ratio. By adding a multiplier in the interpolation chain
the amplification can be compensated
+ Add a HDL parameter for the PPS receiver module :
PPS_RECEIVER_ENABLE. By default the module is disabled.
+ Add the CMOS_OR_LVDS_N and PPS_RECEIVER_ENABLE into the CONFIG
register
+ Define a pps_status read only register, which will be asserted, if the free
running counter reach a certain fixed threshold. (2^28) The register can
be deasserted by an incomming PPS only.
Make sure that the right hand side expression of assignments is not wider
than the target signal. This avoids warnings about implicit truncations.
None of these changes affect the behaviour, just fixes some warnings about
implicit signal truncation.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The AXI specification that the minimum address space size is 4k, make sure
the axi_dmac adheres to this.
Internally the register space is still 2k. This means the upper and lower
2k of the axi4lite register space will map to the same internal registers.
Software must not rely on this and only access the lower 2k to enable
compatibility in case the internal space grows in the future.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Terminate the m_axi_list signal of the data mover instance in the
src_axi_stream module. This avoids a warning about the port being
unconnected.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Currently the axi_dmac_hw.tcl script does not create interfaces if they
are not used in the current configuration. This has the disadvantage that
the ports belonging to these interfaces are not included in the generated
HDL wrapper. Which will generate a fair bunch of warnings when synthesizing
the HDL.
Instead always generate all interfaces, but disable those that are not used
in the current configuration. This will make sure that the ports belonging
to these interfaces are properly tied-off in the generate wrapper HDL.
This reduces the amount of false positive warnings generated and makes it
easier to spot actual issues.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The DMAC currently doesn't support transfers where the length is not a
multiple of the bus width. When generating the wstrb signal we do pretend
though that we do and dynamically generate it based on the LSBs of the
transfer length.
Given that the other parts of the DMA don't support such transfers this is
unnecessary though. So remove it for now and replace it with a constant
expression where wstrb is always fully asserted.
The generated logic for the wstrb signal was quite terrible, so this
improves the timing of the core.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Currently the read side of the src_response interface is not used. This
leads to warnings about signals that have a value assigned but are never
read.
To avoid this just comment out all signals that are related to the
src_response interface for now.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Make sure that the right hand side expression of assignments is not wider
than the target signal. This avoids warnings about implicit truncations.
None of these changes affect the behaviour, just fixes some warnings about
implicit signal truncation.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The external s_axi_{awaddr,araddr} signals that are connect to the core
have their width set according to the specified size of the register map.
If the s_axi_{awaddr,araddr} signal of the core is wider (as it currently
is for many cores) the MSBs of those signals are left unconnected, which
generates a warning.
To avoid this make sure that the signal width matches the declared register
map size.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Currently the util_upack_hw.tcl script does not create interfaces if they
are not used in the current configuration. This has the disadvantage that
the ports belonging to these interfaces are not included in the generated
HDL wrapper. Which will generate a fair bunch of warnings when synthesizing
the HDL.
Instead always generate all interfaces, but disable those that are not used
in the current configuration. This will make sure that the ports belonging
to these interfaces are properly tied-off in the generate wrapper HDL.
This reduces the amount of false positive warnings generated and makes it
easier to spot actual issues.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Currently the util_cpack_hw.tcl script does not create interfaces if they
are not used in the current configuration. This has the disadvantage that
the ports belonging to these interfaces are not included in the generated
HDL wrapper. Which will generate a fair bunch of warnings when synthesizing
the HDL.
Instead always generate all interfaces, but disable those that are not used
in the current configuration. This will make sure that the ports belonging
to these interfaces are properly tied-off in the generate wrapper HDL.
This reduces the amount of false positive warnings generated and makes it
easier to spot actual issues.
While we are at it also use a loop to create the interfaces since they all
follow the same pattern.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Currently the axi_ad9144_hw.tcl script does not create interfaces if they
are not used in the current configuration. This has the disadvantage that
the ports belonging to these interfaces are not included in the generated
HDL wrapper. Which will generate a fair bunch of warnings when synthesizing
the HDL.
Instead always generate all interfaces, but disable those that are not used
in the current configuration. This will make sure that the ports belonging
to these interfaces are properly tied-off in the generate wrapper HDL.
This reduces the amount of false positive warnings generated and makes it
easier to spot actual issues.
While we are at it also use a loop to create the interfaces since they all
follow the same pattern.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The width of a ternary operator expression is the width of the wider of the
two selectable expression. This means the right side expression of the
tx_data assigment is always 256 bits. This generates an implicit truncating
warning if the tx_data signal itself is only 128 bits.
To avoid this slightly reformulate the expression to yield the correct
width depending on the configuration.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The width of a ternary operator expression is the width of the wider of the
two selectable expression. This means depending on the selected DMA_RATIO
the right side expression of the dma_waddr_rel_s assignment can be up to
three bits wider than the dma_waddr_rel_s signal. This generates an
implicit truncation warning.
Slightly reformulate the expression without the use of the ternary operator
to avoid this warning.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The signal is called adc_clk and not adc_clock. None of the designs is
currently using the signal, so this hasn't been an issue other that it
generates a warning.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Comments starting with the word altera are interpreted by the Altera tools
to be synthesis attribute assignments. In this case this is just a generic
comment though which results in a warning that the synthesis attribute is
unknown.
Slightly reword the comment to avoid this. This is not pretty, but better
than having the false positive warning show up in the log.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The PLL frequency must be half of the lane rate and the core clock rate
must be lane rate divided by 40. There is no other option, otherwise things
wont work.
Instead of having to manually specify PLL and core clock frequency derive
them in the transceiver script. This reduces the risk of accidental
misconfiguration.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The clock bridge expects the clock rate to be specified in Hz, but
$m_coreclk_frequency is in MHz. Do the appropriate conversion.
Nothing seems to rely on the clock bridge reporting the correct frequency
at the moment, so this is only a cosmetic change.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The ad_pps_receiver is instantiated at the top of core.
The rcounter is placed into adc/dac_common registers space, at the
address 0x30 (word aligned).
The interrupt mask is placed into adc/dac_common, at the address 0x04
(word aligned). Because the core has an instance of both modules, the
interrupt masks are OR-ed together.
Add a module to receive 1PPS signal from a GPS module. The module has a
free running counter, which runs on the device's interface clock. The
counter value is latched into a register each time when a 1PPS arrives.
An interrupt signal is also generated in every 1PPS.
Add a check to RX register map to confirm that the ILAS memory registers
return the correct values after the ILAS data has been received.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The MSB of the d_count signal is used as a overflow marker to stop the
counter from incrementing in the monitored clock domain. It is not exported
through the register map and truncated when assigned to the up_d_count
signal.
Make the truncation explicit to make it clear that this is not a mistake
and to avoid warnings about implicit truncation.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The generic Altera clock monitor constraints expect the instance to be
called i_clock_mon. Adjust the code accordingly.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
In this particular case the behaviour is the same with non-blocking and
blocking assignments, but that could change if the code is modified in the
future. To avoid any potentially issue due to this consistently use
non-blocking assignments.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The axi_dmac can issue up to FIFO_SIZE read and write requests in parallel.
This is done in order to maximize throughput and compensate for for
latency.
Set the {read,write}IssuingCapability properties accordingly on the AXI
master interfaces. Otherwise qsys might decide to insert bridges that
artificially limit the number of requests, which in turn might affect
performance.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Qsys allows to query to query the clock domain that is associated with a
clock input of a peripheral. This allows to automatically detect whether
the different clocks of the DMAC are asynchronous and CDC logic needs to be
inserted or not.
Auto-detection has the advantages that the configuration parameters don't
need to be set manually and the optional configuration will be choose
automatically. There is also less chance of error of leaving the settings
in a wrong configuration when e.g. the clock domains change.
In case the auto-detection should ever fail configuration options that
provide a manual overwrite are added as well.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Group configuration parameters by function, provide human readable labels
as well as specify the allowed ranges for each parameter.
This prevents accidental misconfiguration and also makes it easier to
inspect (or change) the configuration in the Qsys GUI.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
In this particular case the behaviour is the same with non-blocking and
blocking assignments, but that could change if the code is modified in the
future. To avoid any potentially issue due to this consistently use
non-blocking assignments.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Use the ad_ip_intf_s_axi helper function to create the axi4lite slave
interface for memory mapped peripherals. This slightly reduces the amount
of boilerplate code in the peripheral's *hw.tcl
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The address width of the AXI interface depends on the size of the register
and can differ from peripheral to peripheral. Add a parameter to the
function that allows to specify the address width, this allows to use the
function for more peripherals.
Keep the current value of 16 bits as the default if the parameter is not
specified.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The axi_adxcvr register map only uses a single 4k page, make this explicit.
This will allow for tighter packaging in the limited available total
register map space.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
This partially reverts commit a8ade15173.
Remove the nonsensical Makefile dependencies that got added by accident.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The MSB of the d_count signal is used as a overflow marker to stop the
counter from incrementing in the monitored clock domain. It is not exported
through the register map and truncated when assigned to the up_d_count
signal.
Make the truncation explicit to make it clear that this is not a mistake
and to avoid warnings about implicit truncation.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The generic Altera clock monitor constraints expect the instance to be
called i_clock_mon. Adjust the code accordingly.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
In this particular case the behaviour is the same with non-blocking and
blocking assignments, but that could change if the code is modified in the
future. To avoid any potentially issue due to this consistently use
non-blocking assignments.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The axi_dmac can issue up to FIFO_SIZE read and write requests in parallel.
This is done in order to maximize throughput and compensate for for
latency.
Set the {read,write}IssuingCapability properties accordingly on the AXI
master interfaces. Otherwise qsys might decide to insert bridges that
artificially limit the number of requests, which in turn might affect
performance.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The SYNC signal that gets reported through the status interface should be
the output (second stage) of the synchronizer circuit.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Make sure the core_cfg_transfer_en signal is declared before they are used.
Strictly speaking the current code is correct and synthesis correctly, but
declaring the signals make the intentions of the code more explicit.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Be more standard compliant and assign names to generate for-blocks. This is
required for Altera/Intel support.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Make sure the req_gen_valid and req_gen_ready signals are declared before
they are used. Strictly speaking the current code is correct and synthesis
correctly, but declaring the signals make the intentions of the code more
explicit.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
In some cases, the 'core_ilas_config_data' registers will be infered as
FDRE, instead of FDSE. Therefor a max delay definition, which are using
the S pin as its endpoint, it can become invalid, nonexistent.
Generalize the path, using the register itself as endpoint.
Increase the width of wvalid_counter, should be equal with awlen width.
The wvalid_counter needs to count from zero to the required burst
length. The maximum burst length is 255, so the width of the counter
have to be 8 bits. axi_last_beats will get the last axi burst length.
The fifo will ask for a new data from the DDR, if the current
level is lower than the high threshold. This will prevent overflow.
By deleting the lower threshold, we can avoid ocassional underflows,
when the DAC rate is closer to the max DDRx rate.
All verilog file are using the Verilog-2001 standard to define
and/or declare ports. Definin a port width with a local parameter
is a bad practive, when this standard is used. Some simulators
will crash. Try to avoid it.
Fix the dma_ready mux in top module, and the dma_ready_out reset
logic in axi_dacfifo_wr module. Also, both write and read addresses
of the async CDC fifo (inside the axi_dacfifo_wr) should be reset
before a dma transaction starts.
If the streaming bit is set, after the trigger condition is met
data will be continuosly captured by the DMA. The streaming bit
must be set to 0 to reset triggering.
If the streaming bit is set, after the trigger condition is met,
data will be continuosly captured by the DMA. The streaming bit
must be set to 0 to reset triggering.
In non-streaming mode we want direction changes to be applied immediately.
The current code has a typo and checks the wrong signal. overwrite_data
holds the configured output value of the pin, whereas overwrite_enable
configures whether the pin is in streaming or manual mode.
For correct operation the later signal should be used to decide whether a
direction change should be applied. Otherwise the direction change will
only be applied if the output value of the pin is set to logic high.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
When using non-broadcast access to the GT DRP registers lane filtering is
done on both sides. The ready and data signals are filtered in the in the
axi_adxcvr module and the enable signal is filtered in the util_adxcvr
module. This works fine as long as both sides use the same transceiver IDs.
E.g. channel 0 of the axi_adxcvr module is connected to channel 0 of the
util_adxcvr module.
But this is not always the case. E.g. on the ADRV9371 platform there are
two RX axi_adxcvr modules (RX and RX_OS) connected to the same util_adxcvr.
The first axi_adxcvr uses lane 0 and 1 of the util_adxcvr, the second uses
lane 2 and 3.
Non-broadcast access for the first RX axi_adxcvr module works fine, but
always generates a timeout for the second axi_adxcvr module. This is
because lane 0/1 of the axi_adxcvr module is connected to lane 2/3 of the
util_adxcvr and when ID based filtering is done both can't match at the
same time.
To avoid this perform the filtering for all the signals in the axi_adxcvr
module. This makes sure that the same base ID is used.
This also removes the sel signal from the transceiver interfaces since it
is no longer used on the util_adxcvr side.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Always explicitly specify the signal width for constants to avoid warnings
about signal width mismatch.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The buffer delay should be 0 in the default configuration. The current
value of 0xb must have slipped in by accident.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Use a single standalone counter that counts the number of beats since the
release of the SYNC~ signal, rather than re-using the LMFC counter plus a
dedicated multi-frame counter.
This is slightly simpler in terms of logic and also easier for software to
interpret the data.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
There are currently two sysref related events. One the sysref captured
event which is generated when an external sysref edge has been observed.
The other is the sysref alignment error event which is generated when a
sysref edge is observed that has a different alignment from previously
observed sysref edges.
Capture those events in the register map. This is useful for error
diagnostic. The events are sticky and write-1-to-clear.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The internal LMFC offset signals are in beats, whereas the register map is
in octets. Add the proper alignment padding to the register map to
translate between the two.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
For SYSREF handling there are now three possible modes.
1) Disabled. In this mode the LMFC is generated internally and all external
SYSREF edges are ignored. This mode should be used for subclass 0 when no
external sysref is available.
2) Continuous SYSREF. An external SYSREF signal is required and the LMFC is
aligned to the SYSREF signal. The SYSREF signal is continuously monitored
and if a edge unaligned to the previous edges is detected the LMFC is
re-aligned to the new edge.
3) Oneshot SYSREF. Oneshot SYSREF mode is similar to continuous SYSREF mode
except only the first edge is captured and all further edges are ignored,
re-alignment will not happen.
Both in continuous and oneshot signal at least one external sysref edge is
required before an LMFC is generated. All events that require an LMFC will
be delayed until a SYSREF edge has been captured. This is done to avoid
accidental re-alignment.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
If the output pin is not defined as a clock, some of the Vivado IPI
propagation TCL will error out.
Signed-off-by: Matt Fornero <matt.fornero@mathworks.com>
By adding support for partial avalon transfers (data width < bus width),
valid data set size (DMA transfer length) will be dependent on the DMA bus
width only.
+ avl_write_transfer_done_s is a redundant net
+ specify the net state explicitly on if statements
+ to define the edge of avl_mem_fetch_wr_address signal,
its register and its second sync register should be used
The ad_mem_asym memory read interface has a 3 clock cycle delay, from the
moment of the address change until a valid data arrives on the bus;
because the dac_xfer_out is going to validate the outgoing samples (in conjunction
with the DAC VALID, which is free a running signal), this module will compensate
this delay, to prevent duplicated samples in the beginning of the
transaction.
+ all net names should have a *_s postfix
+ avl_burstcount is a constant 1, no need for an additional
register for it
+ all CDC should have two synchronization register, add
avl_last_beat_req_m2
Lars-Peter Clausen
Rejeesh Kutty
Istvan Csomortani
Adrian Costina
Matthew Fornero
AndreiGrozav