| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
run_testcase() can fail in many ways and just cause runtc to return 1,
at least now it doesn't do so silently.
|
|
|
|
Due to bug #1066, the output of all DR/IR bits changed in
the underlying JTAG decoder.
|
|
As reported in bug #1066, the decoding of IR/DR bits was incorrect.
|
|
|
|
The dcf77 decoder recently started emitting "invalid bit timing" and
"Invalid DCF77 bit number" annotations, and does display bit positions
of the 59th and subsequent pulse within a minute.
|
|
|
|
|
|
|
|
The IR-NEC decoder optionally can handle input streams that still have
the carrier signal. Add a test for this configuration, derived from the
test which uses the signal that had the carrier removed in hardware.
The same data is seen, just at different sample numbers. The content was
verified by means of the "cut -d ' ' -f 2-" and "diff" commands.
|
|
Use a smaller input stream which lacks a samplerate spec. This results
in output that lacks the timing information (motion speed), but step
count and direction are identical to the other existing test.
|
|
The stepper motor decoder can still determine the steps' directions
and count. Missing input sample rates just prevent the decoder from
outputting speed annotations. All other decoder output is identical.
|
|
|
|
Use a smaller input stream which lacks a samplerate spec. This results
in identical decoder output (the existing test did not cover throughput
annotations).
|
|
|
|
The existing test did not check the throughput annotation. Which is why
the decoder output is identical for input streams with and without a
sample rate specification.
|
|
|
|
|
|
Resetting counters depending on the LCD controller's R/S signal yields
numbers that correspond to byte sequences that either "control" hardware
(register access) or carry "data" (display text).
Provide a test case for the rising edge of RESET. Existing cases already
covered the falling edge of RESET which is the default.
|
|
|
|
|
|
|
|
The parallel decoder got adjusted to emit annotations for "items" and
"words" in a unified format (consistent width, leading zeroes). Adjust
the test expectation to match the PD implementation.
|
|
The USB power delivery decoder got adjusted to construct complex
annotations that consisted of several text fragments in a reproducible
order. Adjust the test expectation to match the PD implementation.
|
|
|
|
|
|
|
|
|
|
This fixes (parts of) bug #1019.
|
|
|
|
The CAN decoder got adjusted to synchronize to the input stream's clock.
Catch up in the test expectation.
|
|
The start sample number of onewire network decoder annotations got fixed
in the decoder implementation (used to start one bit late). Catch up in
the test expectation.
|
|
The previous implementation used the period counter instead of the
period's actual start and end sample number. The implementation got
fixed, catch up in the test expectation.
|
|
|
|
|
|
Remove the "exception_samplerate" test for the SPI decoder. Lack of a
samplerate specification only breaks the emission of the throughput
annotations, but still allows decoding the clocked signal.
|
|
|
|
Adapt all tests that require it to use the old behaviour of assuming an
initial pin value of 0 (or whatever was previously hardcoded in the PD).
for now (this will likely see some changes later).
|
|
|
|
Option values are strings, not numbers.
This fixes:
$ ./decoder/pdtest -s jitter
Testcase: jitter/toim4243
Protocol decoder: jitter
Channel clk=3
Channel sig=2
Error: %d format: a number is required, not str
|
|
|
|
|
|
|
|
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
|
|
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
|
|
|
|
Signed-off-by: Karl Palsson <karlp@tweak.net.au>
|
|
Help says -l should list tests. Make it do so.
You can still do "-l uart" to see only uart tests, but -l by itself now
does what you expect.
Signed-off-by: Karl Palsson <karlp@tweak.net.au>
|
