SDSCC/SDS++ Compiler Commands

sdscc | sds++ [hardware_function_options] [system_options]
           [performance_estimation_options]
           [options_passed_through_to_cross_compiler]
           [-sds-pf platform_name] [-sds-pf-info platform_name]
           [-sds-pf-list] [-sds-sys-config configuration_name]
           [-sds-proc processor_name] [-target-os os_name]
           [-debug-xrf] [-debug-xrf-cc compiler]
           [-sds-pf-path path] [-sds-image image_name]
           [-verbose] [-version] [--help] [files]

[-sds-hw function_name source_file [-clkid clock_id_number] 
[-files hls_file_list] [-hls-target boolean_value] [-hls-target-flags "target_options"] [-hls-tcl hls_tcl_directives_file] 
[-shared-aximm] -sds-end]*

[[-perf-funcs function_name_list -perf-root function_name] | 
[-perf-est data_file][-perf-est-hw-only]]

[[-ac function_name:clock_id_number]*[-apm] [-bsp-config-file mss_file] 
[-bsp-config-merge-file mss_file][-disable-ip-cache] [-dm-sharing <0-3>] [-dmclkid clock_id_number] 
[-emulation mode] [-impl-strategy <strategy>] 
[-instrument-stub] [-maxthreads number] [-mno-bitstream][-mno-boot-files] [-rebuild-hardware] 
[-remote-ip-cache cache_directory]
[-synth-strategy <strategy>] [-trace] [-trace-buffer depth] [-trace-no-sw] [-maxjobs <number>] 
[-sdcard <data_directory>][-vpl-ini ini_file] [-xp parameter_value]]

The following command line options are applicable for any sds++ invocation or your display information.

Hardware function options provide a means to consolidate sdscc/sds++ options within a Makefile to simplify command line calls and make minimal modifications to a pre-existing Makefile.

APPSOURCES = add.cpp main.cpp
EXECUTABLE = add.elf
 
CROSS_COMPILE = arm-xilinx-linux-gnueabi-
AR = ${CROSS_COMPILE}ar
LD = ${CROSS_COMPILE}ld
#CC = ${CROSS_COMPILE}g++
PLATFORM = zc702
SDSFLAGS = -sds-pf ${PLATFORM} \
           -sds-hw add add.cpp -clkid 1 -sds-end \
           -dmclkid 2
CC = sds++ ${SDSFLAGS} 

INCDIRS = -I..
LDDIRS =
LDLIBS =
CFLAGS = -Wall -g -c ${INCDIRS}
LDFLAGS = -g ${LDDIRS} ${LDLIBS}
 
SOURCES := $(patsubst %,../%,$(APPSOURCES))
OBJECTS := $(APPSOURCES:.cpp=.o)
 
.PHONY: all
 
all: ${EXECUTABLE}
 
${EXECUTABLE}: ${OBJECTS}
 ${CC} ${OBJECTS} -o $@ ${LDFLAGS}
 
%.o: ../%.cpp
 ${CC} ${CFLAGS} $< 

# synthesis directives
create_clock -period <clock_period>
set_clock_uncertainty 27.0%
config_rtl -reset_level low
source <sdsoc_generated_tcl_directives_file>
# end synthesis directives

A full bitstream compile can take much more time than a software compile, so the sdscc/sds++ (referred to as sds++) applications provide performance estimation options to compute the estimated runtime improvement for a set of hardware function calls.

In the Application Project Settings pane, to invoke the estimator, select the Estimate Performance check box. This enables performance estimation for the current build configuration and builds the project.

Figure: Setting Estimate Performance in Application Project Settings

Estimating the speed-up is a two phase process:

1. The SDSoC environment compiles the hardware functions and generates the system. Instead of synthesizing the system to bitstream, the sds++ computes an estimate of the performance based on estimated latencies for the hardware functions and data transfer time estimates for the callers of hardware functions.
2. In the generated Performance Report, to determine a performance baseline and the performance estimate, select Click Here to run an instrumented version of the software on the target .

See the SDSoC Environment Getting Started Tutorial (UG1028) for a tutorial on how to use the Performance Report.

You can also generate a performance estimate from the command line. As a first pass to gather data about software runtime, use the -perf-funcs option to specify functions to profile and -perf-root to specify the root function encompassing calls to the profiled functions.

The sds++ system compiler then automatically instruments these functions to collect runtime data when the application is run on a board. When you run an instrumented application on the target, the program creates a file on the SD card called swdata.xml, which contains the runtime performance data for the run.

Copy the swdata.xml to the host, and run a build that estimates the performance gain on a per hardware function caller basis and for the top-level function specified by the –perf-root function in the first pass run. Use the –perf-est option to specify swdata.xml as input data for this build.

The SDSoC environment uses the same GNU Arm cross-compiler toolchains included with the Xilinx Software Development Kit (SDK).

The Linaro-based GCC compiler toolchains support the Zynq-7000 SoC and Zynq UltraScale+ MPSoC family devices. This section includes additional information on toolchain usage.

When compiling and linking applications, use only object files and libraries built using the same compiler toolchain and options as those used by the SDSoC environment. All SDSoC provided software libraries and software components (Linux kernel, root filesystem, BSP libraries, and other pre-built libraries) are built with the included toolchains. If you use sdscc/sds++ (referred to as sds++) to compile object files, the tools automatically insert a small number of options. If you invoke the underlying toolchains, you must use the same options.

For example, if you use a different Zynq-7000 SoC floating-point application binary interface (ABI), your binary objects are incompatible and cannot be linked with SDSoC Zynq-7000 binary objects and libraries.

The following table summarizes the sds++ usage of Zynq-7000 SoC toolchains and options. Where options are listed, you need to specify them only if you use the toolchain gcc and g++ commands directly instead of invoking sds++.

The following table summarizes sds++ usage of Zynq UltraScale+ MPSoC Cortex™-A53 toolchains and options. Where options are listed, you only need to specify them if you use the toolchain gcc and g++ commands directly instead of invoking sds++.

The following table summarizes the sds++ usage of Zynq UltraScale+ MPSoC Cortex-R5 toolchains and options. Where options are listed, you need to specify them only if you use the toolchain gcc and g++ commands directly instead of invoking sds++.

For source files that do not contain hardware accelerators or callers but do use NEON intrinsics, you can either compile them directly using the GNU toolchain and link the objects with sds++, or you can add the sds++ command line option -mno-ir for these source files. This option prevents clang-based tools from being invoked to create an intermediate representation (IR) used in analysis. You are programmatically aware that they are not required (such as no accelerators or callers).

For the latter solution, if you are using the SDSoC environment, you can apply the option on a per-file basis by right-clicking the source file, select Properties and go to the Settings dialog box under C/C++ Build Settings > Settings.

Figure: Build Settings