| 08:13 | jnettlet[m]> | topi`_: will look at it in a bit. Just finishing up getting nvme working in u-boot |
| 08:58 | jnettlet[m]> | pci_find_and_bind_driver: Match found: nvme |
| 08:58 | jnettlet[m]> | well I can see the pci device now. But the driver is still not finding it. |
| 08:59 | Ke> | jnettlet[m] I was concerned about nvme having less isolation than sata, how much of this would you consider to make sense |
| 09:00 | Ke> | nvme has only iommu isolation while sata has iommu+ahci controller |
| 09:00 | Ke> | I guess u-boot does not even use iommu? |
| 09:01 | jnettlet[m]> | no |
| 09:02 | Ke> | so your nvme card is to be considered completely trusted |
| 09:14 | jnettlet[m]> | for now. there are patches to support iommu for the pcie layer in uboot being developed |
| 09:32 | jnettlet[m]> | topi`_: do you get i2c errors just before mv_ddr runs as well? ERROR: Status 30 in write transaction |
| 09:37 | jnettlet[m]> | bingo. NVMe: devnum: 0 Vendor: 0x144d Prod: S3ESNX0J607268V |
| 09:38 | jnettlet[m]> | hmmm this isn't right though. Capacity: 40386.6 MB = 39.4 GB (82711872 x 512) |
| 10:09 | jnettlet[m]> | that is more like it. Capacity: 238475.1 MB = 232.8 GB (488397168 x 512) |
| 10:13 | jnettlet[m]> | and I can now read the partitions and filesystem. woot! |
| 10:14 | vpeter> | jnettlet[m]: You probably did something wrong :) |
| 10:15 | jnettlet[m]> | vpeter: I didn't write the driver. Was just posted to the u-boot list. I just fixed it |
| 10:15 | jnettlet[m]> | u-boot loading efi is still crashing though. |
| 12:30 | topi`_> | jnettlet[m]: if I send you the boot logs of both the UBOOT + marvell firmware and the kernel, is that enough to help you detect why cpufreq stopped working? |
| 12:31 | jnettlet[m]> | topi`_: I will just test on my own board to start with. |
| 12:32 | topi`_> | I did not change the kernel underneath. It worked for a while, (after inserting the MSS fw into SPI) but then after I upgraded the mv_ddr, no dice |
| 12:32 | topi`_> | and I upgraded the mv_ddr because my DIMM wasn't getting proper detection... |
| 12:33 | jnettlet[m]> | sure, and we recommend that. It could be that the kernel requires and update as well |
| 12:33 | topi`_> | Linux macchiatobin 4.4.8-mcbin-marvell #1 SMP PREEMPT Mon May 15 18:59:41 UTC 20 |
| 12:33 | topi`_> | still this old one |
| 12:33 | topi`_> | I did not build this, it seems the builder was abuild@armada-8040 |