-
Notifications
You must be signed in to change notification settings - Fork 3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Reliably running host_tests on a u-blox C027 board #992
Comments
The references to the Vector Table issue and pull request are a mistake, please ignore them. |
Hi, |
Just one comment from me, maybe two:
|
Hi, and thanks for responding. Yes, I am testing LPC1768, though only those tests which happen to work on the C027 platform. The .bin files don't appear to be retained; at least, when I ask Windows about the mapped drive it says that 8 kbytes are used and 512 kbytes are free. Since you're travelling I'll go do something else for the remainder of the week. |
Rob,
|
Rob, Ran few tests to check your defect and got this.
|
Interesting. Is that run on a C027? I'm using this ARM-GCC version: 4.9.3 20141119 (release) [ARM/embedded-4_9-branch revision 218278] |
Just to be clear (having re-read the thread it might not be clear): I'm running LPC1768 tests but I am running them on the C027 HW platform, not the LPC1768 HW platform, hence I'm only running those tests which I know should work on the C027 HW platform (which contains an LPC1768 processor). FYI I get the same kind of unstable results if I run any of the C027 tests, it was just that I started out with these LPC1768 tests so thought I'd stick with them as nice, vanilla, tests. The problem, I think, is specific to running tests on the C027 HW platform. |
Yes, it is misunderstanding. So for me on U-BLOX test MBED_A1 works well: So I got my U-BLOX connected:
And I configure test suite to assume my U-BLOX is LPC1768:
And now five MBED_A1 tests in a row:
|
And the same test 20 times LPC1768 tests on U-BLOX C027:
|
And one more time LPC1768 tests on U-BLOX C027:
|
Very odd. Out of interest, are you driving the tests from Linux or Windows? I've just run the tests again with GCC_ARM and with target set up as LPC1768, driven from Windows, and in a loop of 20 I never get a clean run. In my first run of 20, two of them hung, in the next run of 20 six of them hung. With -v output the hang is always immediately after the line "HOST: Reset target..."
Is there any additional debug information I can retrieve for you? Rob |
Hi Rob,
Have you tried to manually flash test and run it multiple times (reset board using 'send bread' command from your terminal) ? |
Aaaaaaaaa! On a hunch, I dug another board out of a desk drawer and that works 100%. There must be something very subtly wrong with the other one. Phew. Sorry for bothering you and thanks for your persistence. |
Przemek Wirkus's push request from here:
#986
...makes it possible to run host_tests on a u-blox C027 board but, even with this change, they are not completely reliable. For instance, running test
MBED_A1
five times gives me:In general I get a timeout about 1/3 of the time. Running with
-v
and looking at a fail case, it says:The first
Unknown property:
is sometimes followed by{
or{{timeout;2.
I've tried adding a 5 secondreset_tout
delay but that doesn't seem to help. Is there a more detailed level of debug I could look at to find out what's going wrong? Watching the C027 board as things run, it's difficult to correlate when the problem occurs with the behaviour of the LEDs, but it is something to do with the end of the copy and the reset I guess. What else could I do to make this reliable?The text was updated successfully, but these errors were encountered: