-
Notifications
You must be signed in to change notification settings - Fork 6
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
Support margin time #21
Conversation
…margin info. This is required for BRCM PCIe switches. Signed-off-by: Sathish Sekar <[email protected]>
…ix/brcm_pcie_switch_settings Fixed the sequence to revert to normal settings before lane margin
Signed-off-by: Sathish Sekar <[email protected]>
…ix/handle-empty-string Replaced the string check to use boolean instead of None
Signed-off-by: Sathish Sekar <[email protected]>
hey @churchmice; instructions on how to fix the DCO check are here https://github.com/opencomputeproject/ocp-diag-pci_lmt/pull/21/checks but first you'll need to rebase your branch on dev; it looks like you started your fork from the
|
hi, @mimir-d , thanks for the detailed instruction. |
that's ok, everyone starts from somewhere. let me know if the above steps dont work |
The following features are enhanced :
Per pcie spec 4.2.18.4
"Software should allow margining to run for at least 10^8 bits margined by the Receiver under test before
switching to the next margin step location (unless the error limit is exceeded)."
The current LMT code does not consider this and the margin is done immediately.
A new -m margin_time option is added to let the lane execute at least margin_time seconds.
When a margin command is sent to the device, it is possible the device will need some time to prepare the margin, in other words, we need to check the step_margin_execution_status is 10 ( margin in progress ) before concluding the margin is in progress.
We run the original code on our products and find the margin results are inconsistent:
After the two fixes, the margin result is consistent.