#asahi-dev · 2022-02-20 17:48
kettenisI'm just trying to prevent U-Boot on the M1 to become too much of a special snowflake#asahi-dev · 2022-02-20 17:42
kettenismarcan: thanks#asahi-dev · 2022-02-20 17:37
kettenisnote that the current u-boot config does not define CMD_EFIDEBUG and doesn't provide a way to save the U-Boot environment#asahi-dev · 2022-02-20 17:34
kettenisjannau: I'm wondering if it is possible to pre-seed the EFI variables in u-boot in a way that makes "bootefi bootmgr" do the right thing#asahi-dev · 2022-02-20 17:25
kettenisif we get something agreed upon and convince the maintainers of the efibootmgr tool to implement it, that would provide a solution for all systems that don't have proper EFI variables support#asahi-dev · 2022-02-20 17:23
kettenisso one thought I had was to have a way for the OS to leave a file behind on the ESP that would then be used to update the EFI variables on the next boot#asahi-dev · 2022-02-20 17:12
kettenistpw_rules: you are free to do that (is rEFInd a thing on arm64)#asahi-dev · 2022-02-20 16:58
kettenisso if things continue to work as normal if you just create a single "install" in the Asahi installer I don't see a problem#asahi-dev · 2022-02-20 16:57
kettenisthat said, most users will not install multiple Linux distros on their machines#asahi-dev · 2022-02-20 16:55
kettenisI'd rather figure out a way to support EFI variables#asahi-dev · 2022-02-20 16:51
kettenisI'm skeptical about this approach; OS installers will be confused by the presence of multiple ESPs#asahi-dev · 2022-02-20 14:36
kettenismarcan: I suppose the easiest way is for you to fork https://github.com/u-boot/u-boot and give me access to that such that I can push my asahi branch to that?#asahi-dev · 2022-02-19 16:18
kettenishmm, did I misunderstand that the CLUSTER_PSTATE register contains the actual power state of the cluster as well as the desired state?#asahi-dev · 2022-02-19 11:39
kettenisjust rtc I think#asahi-dev · 2022-02-19 11:36
kettenisbut maybe it makes sense to have that under github.com/AsahiLinux?#asahi-dev · 2022-02-19 11:35
kettenismarcan: I just created an "asahi" u-boot branch with all the diffs that we still need on top of u-boot master#asahi-dev · 2022-02-17 20:24
kettenisaxboe: looks like osx only; even FreeBSD doesn't have it #asahi-dev · 2022-02-17 11:43
kettenisI think "number of openings" is classic SCSI terminology for the number of commands that can be in flight simultaniously#asahi-dev · 2022-02-17 11:40
kettenisnot optimal, but still plenty fast and it has been rock solid since I added that hack#asahi-dev · 2022-02-17 11:39
kettenisalthough in my case the command actually didn't complete as far as I could tell#asahi-dev · 2022-02-17 11:39
kettenisat some point the completion interrupt for a command never happens#asahi-dev · 2022-02-17 11:38
kettenisthe issue I ran into sounds somewhat similar to what you're seeing#asahi-dev · 2022-02-17 11:38
ketteniswhich effectively means command sumission is serialized#asahi-dev · 2022-02-17 11:37
kettenissven: so on OpenBSD I still set the "number of openings" to 1 to avoid the NVMe from locking up#asahi-dev · 2022-02-17 11:21
kettenisdoes the NVMe core code implement different code paths for MSI and non-MSI?#asahi-dev · 2022-02-17 00:40
kettenisI don't think i've ever lost a filesystem though#asahi-dev · 2022-02-17 00:37
kettenistotally#asahi-dev · 2022-02-16 23:01
kettenisif you pull the plug you pray that fsck can restore your filesystem to a consistent state and accept that you lose data#asahi-dev · 2022-02-16 23:00
kettenisI think that is pretty much the traditional UNIX behaviour:#asahi-dev · 2022-02-16 22:59
kettenisI can confirm that OpenBSD only does a disk cache flush when a filesystem is unmounted, when an explicit cache flush ioctl is issued or when the machine is suspended or powered down#asahi-dev · 2022-02-16 16:50
kettenisbut on these machines mmio to non-PCIE devices is explicitly non-posted#asahi-dev · 2022-02-16 16:49
kettenisI thought the dummy read was necessary in the case writes are posted#asahi-dev · 2022-02-16 16:03
kettenisthe hangs are seen with both AIC1 and AIC2 isn't it?#asahi-dev · 2022-02-16 13:39
kettenis(not implying you actually need one)#asahi-dev · 2022-02-16 13:39
kettenistake another shower!#asahi-dev · 2022-02-16 13:11
kettenisit should be possible to test whether the interrupt is really edge triggered or level triggered at the hardware level#asahi-dev · 2022-02-16 13:10
kettenisprobably#asahi-dev · 2022-02-16 13:03
kettenisnote that aic doesn't seem to have a way to configure level vs. edge#asahi-dev · 2022-02-16 12:58
kettenisuse?#asahi-dev · 2022-02-15 20:20
kettenisvery relaxed#asahi-dev · 2022-02-15 19:16
kettenisI hope the bindings we came up with will be fairly uncontroversial since they largely just use generic stuff#asahi-dev · 2022-02-15 17:25
kettenisright, and it should do so for the cpu nodes as well#asahi-dev · 2022-02-15 17:23
kettenisI mean the FDT not the ADT#asahi-dev · 2022-02-15 17:16
kettenisto what extent does m1n1 trim the device tree?#asahi-dev · 2022-02-13 15:19
kettenismarcan: what is pm_setting used for?#asahi-dev · 2022-02-13 14:31
kettenisI question the wisdom of exposing hardware like that through sysfs, but that's not going to help you ;)#asahi-dev · 2022-02-13 14:25
kettenisI suppose we should avoid exposing regions that have side-effects#asahi-dev · 2022-02-13 11:34
kettenisto set the clock, you just tweak the ffset#asahi-dev · 2022-02-13 11:33
kettenisbasically the clock uses some arbitrary epoch and the offset is used to turn that into an offset from 1970#asahi-dev · 2022-02-13 11:31
kettenisthe SMC keys are discoverable, so if at some point a key to read/write the offset appears the SMC RTC driver can just use that