IRC channel logs

<damo22>what is the ifdef for 64 bit hurd?

<damo22>#if defined(__x86_64__)

<damo22>?

<damo22>i fixed rumpnet on amd64

<damo22>youpi: how would we pass status of pic/ioapic from gnumach to acpi?

<damo22>apparently acpi needs to know which mode is set

<damo22>could we hijack /proc/interrupts?

<damo22>or we need a new mach RPC to expose some property of the interrupt controller?

<damo22>maybe a device_get_status for the irq device?

<jab>so I installed the Hurd on my T60 yesterday. The install had a couple of "glitches". Not too hard to get it to work.

<jab>the "text install" failed, but "psuedo graphical" worked. During the install, on the "enter your password screen", I plugged in my phone charger to the usb port, and "Purism charger blah blah blah text" popped up. Then I could no longer see the text to type in my password.

<jab>So I had to select "go back" and then select "users and passwords".

<jab>Not a huge issue.

<jab>Boot worked just fine...but I do see weird issues with "hd2". I think that may be refeering to the installer CD.

<jab>not sure.

<jab>Also my first "su; apt install i3" failed because it was trying to install packages from the installer CD and not from debian.org

<jab>So I'll fix that later today. Also I had to select "compatibility" mode from the BIOS to boot the machine.

<jab>It could not detect the hardware without the "compatibility" option.

<azert>damo22: wouldn’t a x86 specific flavor and the host_info rpc be the most appropriate place to pass the info about pic vs apic from gnumach to the acpi server?

<sneek>azert, you have 1 message!

<sneek>azert, nexussfan says: i was not able to reproduce it on my amd64 hurd host

<azert>but, ideally, having gnumach just do that as Michael is suggesting in his mail seems the most reasonable solution to me

<azert>if it work it work

<azert>he said “I don't know whether it's possible or advisable to perform an ACPI write operation to the \_PIC mode”

<azert>I think he should just try and see if it works. Surely it

<azert>makes lots of sense

<azert>and I don’t think that the low level acpi library does anything magic or undocumented

<azert>Alternatively, maybe gnumach should expose its build options in a device

<azert>damo22: you are suggesting that the acpi server should deal with this, but it seems weird to me that it does thta

<azert>shpuld be the job of gnumach to take care of interrupt controllers and none else

IRC channel logs

2025-10-09.log