Post Reply 
Timestamp TULIP Virtual Port Feeds
10-24-2024, 07:59 PM
Post: #1
Timestamp TULIP Virtual Port Feeds
Can each line in the various serial feeds be timestamped somehow - console, tracer and HP-il scope? I am doing a lot of correlation between different feeds right now. It is hard to match up where a HPIL transaction happens in the tracer feed. Maybe make this an option that can be turned on / off?

If this is not the best spot for a feature request, please redirect me.

Thanks for the consideration.
Find all posts by this user
Quote this message in a reply
10-25-2024, 07:57 AM
Post: #2
RE: Timestamp TULIP Virtual Port Feeds
Best is to add an issue to the TULIP DevBoard GitHub page. But I can do that if you do not have an account there.

In the console I do not think a timestamp adds any value. For the traces it certainly does. I can add the samplecount (first column in the tracer) to the HP-IL scope, so there is a easy cross reference. There will be some jitter in the counter (HP-IL scope data is generated in the non-critical core) but I estimate this is not too much.
A clock will be a bit of a hassle, and there is not battery backed RTC on the device. A disadvantage of the sample counter is that it resets after every power event.

Regards, Meindert
Find all posts by this user
Quote this message in a reply
10-25-2024, 12:20 PM
Post: #3
RE: Timestamp TULIP Virtual Port Feeds
(10-25-2024 07:57 AM)MeindertKuipers Wrote:  Best is to add an issue to the TULIP DevBoard GitHub page. But I can do that if you do not have an account there.

In the console I do not think a timestamp adds any value. For the traces it certainly does. I can add the samplecount (first column in the tracer) to the HP-IL scope, so there is a easy cross reference. There will be some jitter in the counter (HP-IL scope data is generated in the non-critical core) but I estimate this is not too much.
A clock will be a bit of a hassle, and there is not battery backed RTC on the device. A disadvantage of the sample counter is that it resets after every power event.

Using the samplecount will help immensely, thank you. Right now I am using the cdcident command to bookmark the feeds, and that is w-a-y too granular.

I will add an issue in the github.

-Pat
Find all posts by this user
Quote this message in a reply
10-25-2024, 02:07 PM
Post: #4
RE: Timestamp TULIP Virtual Port Feeds
(10-25-2024 12:20 PM)rocket.scientist Wrote:  
(10-25-2024 07:57 AM)MeindertKuipers Wrote:  Best is to add an issue to the TULIP DevBoard GitHub page. But I can do that if you do not have an account there.

In the console I do not think a timestamp adds any value. For the traces it certainly does. I can add the samplecount (first column in the tracer) to the HP-IL scope, so there is a easy cross reference. There will be some jitter in the counter (HP-IL scope data is generated in the non-critical core) but I estimate this is not too much.
A clock will be a bit of a hassle, and there is not battery backed RTC on the device. A disadvantage of the sample counter is that it resets after every power event.

Using the samplecount will help immensely, thank you. Right now I am using the cdcident command to bookmark the feeds, and that is w-a-y too granular.

I will add an issue in the github.

-Pat
And using the cdcident in the trace wil not help you much. The trace is delayed depending on filters, performance and trace depth.
Got the issue.

Regards, Meindert
Find all posts by this user
Quote this message in a reply
Post Reply 




User(s) browsing this thread: 3 Guest(s)