#
✕ nick:kettenis
#asahi · 2022-03-10 21:13
kettenisyes, we prefer to write code ;)
#asahi · 2022-03-10 14:32
kettenisi'm still not completely sold on the idea to have more than one esp
#asahi · 2022-03-10 14:27
kettenisso maybe we should prefer uefi over efi in the name
#asahi · 2022-03-10 14:26
kettenisDocumentation/arm/uefi.rst
#asahi · 2022-03-10 14:26
kettenismarcan: while we're bikeshedding about the property name, Linux uses a few linux,uefi-mmap-xxxx properties already
#asahi · 2022-03-10 08:01
kettenisnot all linux distros use the same mountpoint, which leads to confused users on this channel ;)
#asahi · 2022-03-10 08:01
kettenissome/many/all linux distros then mount that partition and for that you need a directory as a mount point
#asahi · 2022-03-10 08:00
kettenisyou need an esp partition
#asahi · 2022-03-10 08:00
kettenisyou're probably confused
#asahi · 2022-03-09 23:13
kettenisah right, that might be good enough
#asahi · 2022-03-09 23:07
ketteniswithout efi variable support that is hard
#asahi · 2022-03-08 19:09
kettenishey you're forgetting about rust!
#asahi · 2022-03-08 19:06
kettenisso now we get to test how well you can predict the future with respect to AIC2?
#asahi · 2022-03-08 19:06
kettenisah
#asahi · 2022-03-08 19:05
kettenisso M1 Ultra is 2 M1 Pro dies we reckon?
#asahi · 2022-03-08 07:56
kettenis(glass console == framebuffer)
#asahi · 2022-03-08 07:56
kettenishowever, the current u-boot on the asahi branch has this disabled and will always use the glass console
#asahi · 2022-03-08 07:55
kettenisunplug keyboard -> serial console
#asahi · 2022-03-08 07:55
kettenisplug in a keyboard -> glass console
#asahi · 2022-03-08 07:55
kettenistpw_rules: yes, it works for keyboards on usb-c
#asahi · 2022-03-07 20:44
kettenisthe second is m1n1 itself, which is dependent on where iBoot loads m1n1 isn't it?
#asahi · 2022-03-07 20:44
kettenisthe first is for the FDT and allocated by m1n1
#asahi · 2022-03-07 20:44
kettenisso m1n1 has two reservations
#asahi · 2022-03-07 20:35
kettenisu-boot totally lacks the infrastructure to do this in a sensible way
#asahi · 2022-03-07 20:35
ketteniskeeping all those addresses at the top would probably help
#asahi · 2022-03-07 20:28
kettenisso yes, it seems that we need a bit more smarts in the code that sets up the address variables
#asahi · 2022-03-07 20:27
kettenisyes, and m1n1 needs to mark those as reserved somehow
#asahi · 2022-03-07 20:25
kettenisah, maybe not
#asahi · 2022-03-07 20:19
kettenisat least that is my understanding
#asahi · 2022-03-07 20:19
kettenisah, but if thise cover part of normal memory, those should be covered by reserved-memory as well
#asahi · 2022-03-07 20:17
kettenisbut we don't have any reserved-memory nodes, at least not yet
#asahi · 2022-03-07 16:52
kettenisno the efi loader uses $loadaddr by default
#asahi · 2022-03-07 16:48
kettenisor can you use loadaddr
#asahi · 2022-03-07 16:47
kettenisdo you have to use kernel_addr_r?
#asahi · 2022-03-07 16:47
kettenisor it overlaps the device tree
#asahi · 2022-03-07 16:47
ketteniswhat I fear is that somehow that region ends up overlapping another region
#asahi · 2022-03-07 16:46
kettenisthe code jannau added reserves 128M - 2M for the kernel, su there should be plenty of space
#asahi · 2022-03-07 14:54
kettenisstill curious what you did ;)
#asahi · 2022-03-07 14:52
kettenisj'ey: is this reproducable?
#asahi · 2022-03-07 14:49
kettenisif this is an t8103 system, yes
#asahi · 2022-03-07 14:47
kettenisexcept that some variables have an explicit 0x and others don't ;)
#asahi · 2022-03-07 14:46
kettenisI think you did ;)
#asahi · 2022-03-07 14:42
kettenisthe dynamic assignment of the address variables in U-Boot might not be 100% fool-proof
#asahi · 2022-03-07 14:42
kettenisj`ey: I've seen that at some point, but not recently
#asahi · 2022-03-07 14:31
kettenisbut I'm not sure that functionality is exposed through extlinux.conf
#asahi · 2022-03-07 14:30
kettenisnote that you don't have to use grub; U-Boot can in principle load your kernel and boot it using the EFISTUB
#asahi · 2022-03-07 14:28
kettenisso even s2idle will be better with EFI
#asahi · 2022-03-07 14:28
kettenisdeep sleep will probably already require PSCI
#asahi · 2022-03-07 14:27
kettenisultimately that may mean no suspend/resume support if you don't boot using EFISTUB
#asahi · 2022-03-07 14:26
kettenisand the proposed EFI PSCI conduit

plus ›