#asahi-dev · 2021-12-29 22:54
kettenistpw_rules: thanks#asahi-dev · 2021-12-29 12:54
kettenislooks like nvme_pci_configure_admin_queue() does the disable#asahi-dev · 2021-12-29 12:42
kettenisthat doesn't reset the admin queue registers but as long as those get set up correctly before CC.EN is set to 1 again, I don't see how this could mess things up#asahi-dev · 2021-12-29 12:37
kettenispushed a change to clear CC.EN in U-Boot which should do a controller level reset#asahi-dev · 2021-12-29 12:17
kettenisthe first thing the OpenBSD driver does is disabling the controller, so that's probably why I haven't been bitten by this#asahi-dev · 2021-12-29 12:16
kettenisCC.EN is still 1#asahi-dev · 2021-12-29 12:16
kettenishmm, looks like it defenitely doesn't reset the NVMe bits#asahi-dev · 2021-12-29 12:06
kettenissven: the values programmed into the admin queue registers (AQA, ASQ, ACQ) apparently are supposed to survive a reset#asahi-dev · 2021-12-25 11:29
kettenisthis makes no sense to me#asahi-dev · 2021-12-25 11:29
kettenisthe corellium code scribbless some seemingly random bits into the SPMI PMU#asahi-dev · 2021-12-25 11:28
kettenisbtw, related to the SMC, we need to figure out how power down actually works on these machines#asahi-dev · 2021-12-25 11:19
kettenisheh, if you've already figured out how it works...#asahi-dev · 2021-12-25 11:00
kettenismight be worth pushing what you have forward first, to get an idea how the driver maintainers like this apple-specific way of handling firmware...#asahi-dev · 2021-12-23 14:48
kettenisheh, we had a similar issue (although I'm not sure we actually overflowed the buffer)#asahi-dev · 2021-12-23 14:47
ketteniscool#asahi-dev · 2021-12-23 14:46
kettenisreplacing it would require more string parsing code in the kernel, which we don't really like#asahi-dev · 2021-12-23 14:46
kettenisadmittedly the OpenBSD code is a bit dumb and blindly appends a macaddr= line if a local-mac-address property is found in the device tree#asahi-dev · 2021-12-23 14:44
kettenisah, ok, thanks#asahi-dev · 2021-12-23 14:43
kettenisthat'll break OpenBSD#asahi-dev · 2021-12-23 14:43
kettenismarcan: are you really appending a macaddr= line now to the nvram .txt files?#asahi-dev · 2021-12-22 19:38
kettenisyes#asahi-dev · 2021-12-22 19:35
kettenisotherwise it should be a RESEND#asahi-dev · 2021-12-22 19:35
kettenisif you make changes it should be v2#asahi-dev · 2021-12-22 18:46
kettenisI recommend using git send-mail to send patches#asahi-dev · 2021-12-22 18:46
kettenisalso, it seems your mailer corrupts the diff#asahi-dev · 2021-12-22 08:54
kettenismy u-boot currently does power on the wifi, but this probably is something that Linux should do#asahi-dev · 2021-12-22 08:53
kettenisarch/riscv/boot/dts/sifive/fu740-c000.dtsi has a pwren-gpios property next to the reset-gpios property for this purpose#asahi-dev · 2021-12-22 08:52
kettenisto me it looks similar to how the hifive unmatched riscv64 board behaves#asahi-dev · 2021-12-22 08:50
kettenisat least that is what it looks like to me#asahi-dev · 2021-12-22 08:49
kettenisright#asahi-dev · 2021-12-22 08:49
ketteniswifi/bluetooth on all machines, but also SD card on the new machines I believe#asahi-dev · 2021-12-22 08:48
kettenismaz: the SMC controls the GPIO that provides power to some of the PCIe devices that are connected to the RC#asahi-dev · 2021-12-21 20:16
ketteniswould be great if you could get that issue fixed#asahi-dev · 2021-12-21 20:14
kettenisyes, that looks familliar#asahi-dev · 2021-12-21 19:53
kettenisthe mail I remember was about a fix that got rejected#asahi-dev · 2021-12-21 19:52
kettenisbut this does need to be fixed upstream#asahi-dev · 2021-12-21 19:52
kettenismaybe#asahi-dev · 2021-12-21 19:51
kettenisyeah sorry, can't remember the details#asahi-dev · 2021-12-21 19:30
kettenisthat's how u-boot currently handles it#asahi-dev · 2021-12-21 19:30
kettenisj`ey: probably the pcie host bridge driver will have to do that#asahi-dev · 2021-12-21 19:25
kettenisthere was a post on the mailing list somewhere this year#asahi-dev · 2021-12-21 19:20
kettenistpw_rules: a know issue#asahi-dev · 2021-12-21 19:20
kettenisthe keys are discoverable#asahi-dev · 2021-12-21 19:08
kettenisso some good RE-ing of how the "gP??" SMC key really works would not go amiss#asahi-dev · 2021-12-21 19:07
kettenisthat obviously needs some thought#asahi-dev · 2021-12-21 19:07
kettenison the SMC node, corellium has this ugly "gpio-13: 00800000" property#asahi-dev · 2021-12-21 19:06
kettenisthere is a precedent for this for the hifive unmatched PCIe bridge#asahi-dev · 2021-12-21 19:05
kettenisI currently use a "pwren-gpios" property on the PCIe host bridge port#asahi-dev · 2021-12-21 19:04
kettenismarcan: any thoughts yet on the device tree bindings for the SMC GPIOs?#asahi-dev · 2021-12-21 09:23
kettenistpw_rules: disable PSCI_RESET makes sure CONFIG_EFI_HAVE_RUNTIME_RESET isn't set, so this should make u-boot stop advertising EFI runtime reset support