#asahi · 2021-05-31 09:47
kettenisI believe maz' pcie code uses the msi-ranges property that I added#asahi · 2021-05-31 09:45
kettenisu-boot needs the "apple,dart-m1" compatible on the DART nodes for now#asahi · 2021-05-31 09:44
kettenisand maybe touch up the iommu references#asahi · 2021-05-31 09:43
kettenisignore the "apple,pcie-m1" compatible and the extra interrupts#asahi · 2021-05-31 09:41
kettenisif you're using the latest apple-m1-m1n1-nvme branch you can probably just use the pcie node from that#asahi · 2021-05-31 09:39
kettenisbut that tree should be usable as well#asahi · 2021-05-31 09:39
kettenismy u-boot currently has a bit of a frankendevicetree#asahi · 2021-05-31 09:38
kettenissvenpeter: I sent an update of the pcie binding series last night#asahi · 2021-05-29 17:01
kettenisvendor "Broadcom", unknown product 0x5f69 (class network subclass miscellaneous, rev 0x03) at pci1 dev 0 function 1 not configured#asahi · 2021-05-29 17:00
kettenisthere is a function 1 next to the wifi that's probably bluetooth#asahi · 2021-05-29 17:00
kettenis(probably shouldn't call the machine m1n1)#asahi · 2021-05-29 16:59
kettenisthe new 10G m1n1 has been reported to have an aquantia chip instead of the broadcom#asahi · 2021-05-29 16:57
kettenisbge0 at pci3 dev 0 function 0 "Broadcom BCM57762" rev 0x00, BCM57766 A0 (0x57766000): msi, address 00:10:18:00:00:00#asahi · 2021-05-29 16:57
kettenisxhci0 at pci2 dev 0 function 0 "Fresco Logic FL1100 xHCI" rev 0x10: msi, xHCI 1.0#asahi · 2021-05-29 16:56
kettenisbwfm0 at pci1 dev 0 function 0 "Broadcom BCM4378" rev 0x03: msi#asahi · 2021-05-29 16:56
ketteniswifi is indeed broadcom#asahi · 2021-05-29 16:53
kettenisbasically, as soon as you have working pcie, you have type-A usb and ethernet#asahi · 2021-05-29 12:15
kettenisanyway, I'm getting on my bike for a bit#asahi · 2021-05-29 12:14
kettenisdon't cross the...#asahi · 2021-05-29 12:13
kettenisthat certainly doesn't help...#asahi · 2021-05-29 11:55
kettenisit just does the magic to wake up the controller#asahi · 2021-05-29 11:54
kettenisthe nvme support on that branch is still incomplete#asahi · 2021-05-29 11:53
kettenisthat now has i2c and dwc3 support#asahi · 2021-05-29 11:53
kettenissven: https://github.com/kettenis/u-boot/tree/apple-m1-m1n1-nvme#asahi · 2021-05-29 11:36
kettenismaybe I should do that now and worry about trying DART bypass later..#asahi · 2021-05-29 11:34
kettenisI didn't push the u-boot changes yet#asahi · 2021-05-29 11:33
kettenison the openbsd side I just have a dummy driver for the glue node#asahi · 2021-05-29 11:33
kettenisbut otherwise I'm using the standard u-boot dwc3 code#asahi · 2021-05-29 11:32
kettenisand I have a diff to recognise this particular variant of the DesigWare USB3 core#asahi · 2021-05-29 11:32
kettenisI have some hacks in u-boot to trigger that set up a DMA offset for the DART#asahi · 2021-05-29 11:30
kettenisyes#asahi · 2021-05-29 11:21
kettenisbut I may make some noise about something like a "technology preview" when I've sorted out nvme support in u-boot#asahi · 2021-05-29 11:20
kettenisNot yet listing Apple M1 as a "supported" platform for OpenBSD as there are some painful device tree binding bumps ahead#asahi · 2021-05-29 11:18
kettenisno, I'm hanging out here, collaborating with the other AsahiLinix folks#asahi · 2021-05-29 11:17
kettenisthere is no specific project#asahi · 2021-05-29 11:17
kettenishttp://www.openbsd.org/arm64.html#asahi · 2021-05-29 11:16
kettenisI thought it was RHEL that was hostile to non-server arm64#asahi · 2021-05-28 19:33
kettenisguess I need to repartition the disk now#asahi · 2021-05-28 19:33
kettenisso I can now access the NVMe disk#asahi · 2021-05-28 19:33
kettenisand I just fixed the last critical bug#asahi · 2021-05-28 19:32
kettenisDavid Gwynne and Jonathan Matthew wrote a driver for the NVMe controller for OpenBSD#asahi · 2021-05-27 17:03
kettenismarcan: the snapshot stuff might be because the kernel panic'ed#asahi · 2021-05-27 15:52
kettenisyou're lucky that the GPU uses main memory instead f VRAM out on the PCI bus#asahi · 2021-05-27 14:58
kettenisgood enough I'd say ;)#asahi · 2021-05-27 14:54
kettenisit may make sense to only do the pcie_init later when booting a "kernel"?#asahi · 2021-05-27 14:53
kettenisso I removed that code from u-boot#asahi · 2021-05-27 14:53
kettenisI had u-boot hang if I repeated some of the initialization steps that m1n1 now does#asahi · 2021-05-27 14:47
kettenismarcan: the pcie initialization code that's now in m1n1 might make macos unhappy#asahi · 2021-05-25 13:48
kettenisI'll see if I can come up with a diff that does just the clocks and usb_phy_bringup()#asahi · 2021-05-25 13:39
kettenisthe other bits don't seem to hurt ;)