Project

General

Profile

Activity

From 05/30/2015 to 06/28/2015

06/03/2015

AL 09:15 AM Software Development: RE: PCI-e Intel - System crash
Hi Mike,
I'm using the pre-compiled reference image from your PCI Hard IP page. I didn't try to compile the project myself (although I suppose the result will be the same if I do so).
Thanks for all the support.
Alex
Alexandre Lopes
MW 08:44 AM Software Development: RE: PCI-e Intel - System crash
Hi Alex,
Sorry about this. My guess is that the FPGA project we provided is either not routing the PCIe interrupt to the same spot in the HPS that the device tree is configured for or perhaps not at all. The QSYS project would show ...
Michael Williamson
AL 04:45 AM Software Development: RE: PCI-e Intel - System crash
Okay, a couple more informations:
The 3.12 image provided by Critical Link on the PCI Hard IP page seems to have been compiled without CONFIG_PCI_MSI=y, since when I try to load the intel driver with MSI interrupts compiled in I get
Alexandre Lopes

06/02/2015

AL 12:43 PM Software Development: RE: PCI-e Intel - System crash
Comparing both outputs of lspci, it seems that in the MitySoM case, MSIs are disabled: Alexandre Lopes
AL 12:36 PM Software Development: RE: PCI-e Intel - System crash
Hi Mike,
So I have tested both cards on a x86 PC with Kernel 3.12 and 3.18.9 and everything works as it should (using the same drivers, i.e. E1000E and XHCI).
I have also downloaded the pre-compiled image (PCIe Root Port with MSI) ...
Alexandre Lopes

06/01/2015

AL 10:34 AM Software Development: RE: PCI-e Intel - System crash
Hi Mike,
Thanks for your support.
The boot pci-e related messages are (using Critical Link's images):
Alexandre Lopes
MW 09:57 AM Software Development: RE: PCI-e Intel - System crash
Yeah,
I am wondering if this error message might be an issue:
Michael Williamson
MW 09:53 AM Software Development: RE: PCI-e Intel - System crash
Hi Alex,
On the DevKit, we have found that the PCIe card edge is a little shakey without a mechanical mount (e.g., if the card tips forward and backward the PCIe signalling doesn't work well).
Are you running the same Linux version...
Michael Williamson
AL 09:15 AM Software Development: RE: PCI-e Intel - System crash
I have tested a NEC chipset based USB 3.0 PCI-e controller card and I also can't manage to get everything to work.
Again I have tested the official 3.12 image from Critical Link (plus official DTB and FPGA image) and have tested with my...
Alexandre Lopes
 

Also available in: Atom