The 2^18 dma address width with a 64 dma data width will result a FIFO, what will be implemented by 512 RAMB36 cells.
This is a the maximum BRAM FIFO depth in case of the VC707.
The 2^18 dma address width with a 64 dma data width will result a FIFO, what will be implemented by 512 RAMB36 cells.
This is a the maximum BRAM FIFO depth in case of the VC707.
The paths from the HDMI interface registers to the IO pads are
unconstrained. This means the P&R can in theory put the register anywhere
which could lead to stability issues on the interface, depending on what
else is in the fabric. To get predictable delays for the register to IO pad
path place the register into the IOB section.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The HP memory ports on ZYNQ are AXI3. The AXI-DMAC supports both native AXI3
and AXI4, by configuring it for AXI3 there is no need for a protocol
converter inside the interconnect, that connects the DMAC to the HP port.
In addition to that also set the data width for the DMAC on the HP port side
to 64 so there is no need for a memory width converter in the interconnect.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The toplevel input/output signal names are lower case, but the signals
connected to the system_wrapper are upper case. Since verilog is case
sensitive this leaves the toplevel input/output signals unconnected. Fix
this by using lower case names everywhere.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The toplevel input/output signal names are lower case, but the signals
connected to the system_wrapper are upper case. Since verilog is case
sensitive this leaves the toplevel input/output signals unconnected. Fix
this by using lower case names everywhere.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The toplevel input/output signal names are lower case, but the signals
connected to the system_wrapper are upper case. Since verilog is case
sensitive this leaves the toplevel input/output signals unconnected. Fix
this by using lower case names everywhere.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The constraints processing order changed to "late" instead of "early", in order for all the clocks in the system to be already created when the IP constraints are applied
While the reset for the memory mapped AXI master is synchronous to some
clock it is not necessarily synchronous to the clock used for that
interface. So always generate a local reset signal to avoid problems that
could result from this.
While we are at it also update the code to only generate a local reset if
the interface is asynchronous to the register map, otherwise use the
register map reset.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Ignore the timing path from the current DMA address to the register map,
this is just a debug signal at the moment.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
If the internal FIFO is larger than one block ram there will be multiple
BRAMs called ram_reg[0], ram_reg[1]. Modify the BRAM constraint rule so that
it matches these as well.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Set the ASYNC_REG property on the bit synchronizer CDC control regs. This
hint to Vivado that the registers are used for CDC purposes.
Also use -datapath_only for the set_max_delay constraints on the CDC data
path to remove the hold time requirement.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Currently the axi_hdmi_tx core constraints marks all its clocks asynchronous
to all other clocks in the system. This is a bit unfortunate as these
constraints are not restricted to the axi_hdmi_tx, but affect all cores in
the system, some of which might actually have timing constraints on CDC
paths.
The proper way to fix this is to add constraints for the axi_hdmi_tx core
CDC paths. For now only mark the interface clock asynchronous to the HDMI
clock, as this is easy to do and an improvement over the current situation,
as other cores are no longer affected.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
On ZYNQ the HP interconnects have a AXI3 interface. The DMA controller
supports both AXI4 and AXI3. By switching to AXI3 there is no need to create
a protocol converter between the DMA and the HP port.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The AXI master interface and the register map AXI slave interface use the
same clock. No need to mark the interfaces as asynchronous. This removes the
need for CDC logic on those paths.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
The src_response_fifo has been removed from the design, but we still need to
assert the ready and empty control signals for things to work properly.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
We really only want to apply the CDC constraints if the clocks are actually
asynchronous. Unfortunately we can't use if ... inside a xdc script. But we
can use expr which has support for a ? b : c if-like expression. We can use
that to create helper variables that contains valid clock when the clock
domains are asynchronous or {} if they are not. Passing {} as
set_false_path/set_max_delay as either the source or destination will cause
it to abort and no constraints will be added.
Also add -quiet parameters to avoid generating warning if the constraints
could not be added.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
All the FPGA internal control signals are active high, using a active low
reset inserts a extra invert LUT. By using a active high reset we can avoid
that.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Adrian Costina
Istvan Csomortani
Lars-Peter Clausen
Rejeesh Kutty