Rename the variable name used by SOF script with the one expected by the
Zephyr build system for simplicity and consistency.
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
Build for MTL needs other compiler xt-clang, so the COMPILER is
set by default to xt-xcc and changed for MTL in set_xtensa_params.sh
where also core and tools version is defined.
Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
It's not clear why `LDFLAGS=-Wl,-LE` was added in
commit 10d0b3b5e1 ("Scripts: Add xt-run build target for
rebuild-testbench.sh"). Maybe it was supposed to be `-Wl,-EL`?
Either way this flag has always been interpreted as adding the directory
`./E/` to the search library path, which obviously never had any
effect; proof below. So it can safely be removed.
Here's the proof:
```
mv ~/XCC/install/builds/RG-2017.8-linux/cavs2x_LX6HiFi3_2017_8/xtensa-elf/lib/libm.a .
./scripts/rebuild-testbench.sh -p tgl
# Fails as expected
# => XCC/install/tools/RG-2017.8-linux/XtensaTools/bin/xt-ld: cannot find -lm
cmake --build tools/testbench/build_xt_testbench/ -- testbench
# Same again
# => XCC/install/tools/RG-2017.8-linux/XtensaTools/bin/xt-ld: cannot find -lm
mkdir tools/testbench/build_xt_testbench/E/
cp libm.a tools/testbench/build_xt_testbench/E/
cmake --build tools/testbench/build_xt_testbench/ -- testbench
# => now compiles!
```
Remember: the best way to test software is always to break it.
Don't forget to fix XCC:
```
mv libm.a ~/XCC/install/builds/RG-2017.8-linux/cavs2x_LX6HiFi3_2017_8/xtensa-elf/lib/
```
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
This is required to investigate build failures. Because of the way the
build is split, a failure in one component can become totally drown in
the long build logs of another component.
Example: with -j2 the `parser_ep` build failure below is completely
drown by the long (and successful!) build logs of `sof_ep`:
```
./scripts/rebuild-testbench.sh -j 1 -p tgl
sof/tools/testbench/build_xt_testbench/sof_parser/build/include/alsa/
sound/asoc.h:196: error: expected specifier-qualifier-list before ‘__le32’
```
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
This patch adds to rebuild-testbench option -x <platform> that
can be used to build testbench for xt-run execution. The enhanced
script reuses native testbench build but with CC, LD, LDFLAGS,
etc. defines to use the xt-xcc compiler for build.
Currently TGL (HiFi3) is the only supported platform. More will
be added later.
Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
This makes it possible to switch to Ninja with a single line change
which can be useful to test build changes and issues.
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
"make + make install" duplicates many steps, probably because of the
fatal combination of build timestamps + source hash, see PR #3353 for a
similar example.
A single, combined "make install" generates the exact same binary
outputs 30% faster and prints 30% shorter build logs.
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
Default is verbose enough. VERBOSE=1 can still be used later at make
time.
Also change nproc to nproc --all because we don't care about offline
processors.
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
Currently, the testbench is built using the script named
'host-build-all.sh' under scripts directory. But the name of
the script, it's git history doesn't refer to what the script
is currently doing now.
The newly written script adds option to build the testbench.
This script was written in mind to remove the old script
used to build the testbench (i.e host-build-all.sh).
The script also adds an option to build the testbench with
AFL instrumentation. AFL fuzzer works well when the code
it's trying to fuzz is instrumented properly. The instrumentation
helps the fuzzer in generating unique inputs which cover
different paths in the code graph. When the option is set,
testbench is built using a compiler AFL provides, which does
the instrumentation work.
Signed-off-by: Mohana Datta Yelugoti <ymdatta.work@gmail.com>
Marc Herbert