<pch>
who again is making the potential radeon accelerator driver?
<pch>
I have identified a deviance in my own card that might be confusing during testing
<pch>
the card does not correctly report its name, and completely ignores OS mem configuration
<pch>
explains the issue i was having with my 4k panel, but this is still potentially inconvenient for drivertesting
<pch>
from testing some vbios modifications I've also identified that the vmem ignores most vbios config as well
Gaspare has joined #haiku
Gaspare has quit [Ping timeout: 480 seconds]
bitigchi has quit [Ping timeout: 480 seconds]
andreasdr has quit [Remote host closed the connection]
tgschultz has quit [Quit: Going offline, see ya! (www.adiirc.com)]
tgschultz has joined #haiku
maddie has joined #haiku
<maddie>
I'm trying to compile haiku from haiku 64-bit in a VM. I followed the docs on haiku-os.org but when I run the jam command it fails with "zstd.h not found". Poking around the system there is only a zstd.h in /boot/home/queries but nowhere else. The zstd package is installed. I've tried with beta 3 and a nightly image I grabbed yesterday.
Maturion has quit [Remote host closed the connection]
<x512[m]>
pch: "who again is making the potential radeon accelerator driver?" Me?
<x512[m]>
Also Kallisti5 made common and modesetting part some years ago.
<Diver>
maddie: pkgman install zstd_devel
<Diver>
it should in in requires tho. which guide did you use?
<maddie>
Diver: thanks, I missed it. I saw the part that said "Nightly images (or the latest stable release) of Haiku should contain all of the software needed to build Haiku in Haiku." and I tried searching for a devel zstd package in HaikuDepot but didn't try a pkgman command. Reading comprehension failure.
<Diver>
maybe we should adjust that section a bit
bitigchi has joined #haiku
bitigchi has quit [Ping timeout: 480 seconds]
B2IA has quit [Quit: Vision[]: i've been blurred!]
B2IA has joined #haiku
jjido has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
Anarchos has joined #haiku
tgschultz has quit [Quit: Going offline, see ya! (www.adiirc.com)]
tgschultz has joined #haiku
nullman has joined #haiku
x10z has joined #haiku
rennj has quit [Ping timeout: 480 seconds]
rennj has joined #haiku
x10z_ has joined #haiku
WoC has joined #haiku
x10z has quit [Ping timeout: 480 seconds]
freakazoid333 has joined #haiku
x10z_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
AlwaysLivid has joined #haiku
<AlwaysLivid>
hi, i just want to be sure about a step in the haiku compilation process
<AlwaysLivid>
when i compile anyboot images, i'll still produce raw disk images, but they will be turned into an .iso later on. is that correct?
bitigchi has joined #haiku
boistordu_ex has quit [Remote host closed the connection]
<Diver>
AlwaysLivid: anyboot images are iso images
<AlwaysLivid>
i know
<AlwaysLivid>
i'm talking about what happens exactly and in what order - as in, code gets compiled and linked, a basic set of packages is compiled, all of that gets stuffed into a raw disk image alongside with other things that get put in other raw disk images, and then all of that is turned into an anyboot/.iso. is that correct? i'm trying to improve documentation here and explain a process in the simple way that i just brought up
<waddlesplash>
I think we may write to the anyboot directly
<waddlesplash>
but it's been a while since I looked into this. honestly our image building process is quite complicated and somewhat duplicated
<PulkoMandy>
yes I think we populate the anyboot image directly. It would be a waste of time and space to generate a raw image and then copy it over to an anyboot image which has just a few kilobytes of extra stuff at the start
<PulkoMandy>
so the process is basically 1) create an image of the right type (raw, anyboot with a partition table, or reuse an existing device if you are building directly to an existing partition) 2) format as BFS, 3) put the files in it using bfs_shell
<PulkoMandy>
the exception to this process is if you're building from haiku, and directly to an existing mounted bfs partition (using jam @install), in that case the files are just copied to that partition
<AlwaysLivid>
jam @Install is a thing?
<AlwaysLivid>
i was also about to ask that because i wanted to make improvements to aboutsystem
<x512[m]>
I wonder it works with packagefs.
mmu_man has quit [Ping timeout: 480 seconds]
x10z has joined #haiku
AlienSoldier has joined #haiku
<PulkoMandy>
what would be the problem? It installs to a non-booted partition, so it just copies the hpkg files there
<PulkoMandy>
works without problems, that's how I do all my testing when I need to reboot to test my change
<PulkoMandy>
it doesn't update your running install, or at least that is not officially supported
<PulkoMandy>
there is a separate jam @update but that never worked properly, even before packagefs
<PulkoMandy>
if you only want to improve AboutSystem, the easiest way is to run Haiku, and compile only AboutSystem (jam -q AboutSystem) and run it directly from the build directory
<PulkoMandy>
but indeed, all these tips should be documented in docs/develop
HaikuUser has joined #haiku
<waddlesplash>
@update could work now, it should just invoke pkgman
<waddlesplash>
to install the newly built packages
<PulkoMandy>
yes
<AlwaysLivid>
is that documented anywhere?
Anarchos has quit [Remote host closed the connection]
<PulkoMandy>
probably old mailing list posts/commit messages from when it was introduced? And maybe in the jamfile where the build profiles are defined
<PulkoMandy>
(which I guess counts mostly as a no)
x10z has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
maddie has quit [Remote host closed the connection]
madeline_ has joined #haiku
madeline_ is now known as maddie
<maddie>
Building haiku 64 on debian 11, I got "failed BuildFloppyBootImage1", installing bc fixed the issue and I'm getting further in the compile now. bc isn't listed on the pre-reqs page for apt distros, I'm not sure if it's part of a standard ubuntu install but it isn't on debian
<Diver>
probably bc was part of ubuntu before and now it's not?
HaikuUser has quit [Quit: Vision[]: i've been blurred!]
<maddie>
I'm just mentioning it in case the docs need to be updated, I'm not sure the proper channels to bring that up so I brought it up here. bc may still be part of ubuntu idk, I've just found it isn't in debian 11
jjido has joined #haiku
<maddie>
my build just finished successfully, bc was the only missing piece
<nekobot>
[haiku/haiku] c52bbf73b566 - ArchitectureRules: Adjustments to Werror flags for GCC 11.
<nekobot>
[haiku/haiku] a81d165008c8 - Update build-packages for GCC 11 upgrade.
<waddlesplash>
well, that was probably the most annoying gcc upgrade we've ever done
<waddlesplash>
no idea how much time I spent on that last and now this month, but it was al ot
<waddlesplash>
have fun with the buildtools rebuilds y'all
rennj has quit [Quit: Leaving]
<AlwaysLivid>
i'm reproducing a crash by being on the latest nightly, using qemu/virt-manager and git cloning the haiku repo and the buildtools repo at the same time
<waddlesplash>
what, a kdl?
<AlwaysLivid>
yes
<AlwaysLivid>
not sure if i should report that because it sounds completely arbitrary in nature
<AlwaysLivid>
"oh just run git clone twice you'll definitely get a crash"
tgschultz has quit [Quit: Going offline, see ya! (www.adiirc.com)]
tgschultz has joined #haiku
<AlwaysLivid>
seems filesystem relevant, definitely reporting it
<waddlesplash>
kdl tickets are useless without thenm
jesse_blue has joined #haiku
<AlwaysLivid>
wasnt sure whether there was some sort of text file so i expected input tbh
<AlwaysLivid>
because people in $other_communities get annoyed when screenshots are attached cuz text isn't c&p-able
<waddlesplash>
well yeah we want text ideally
<waddlesplash>
but kdls only get text if they're continuable
<waddlesplash>
and generally even attempting to continue from a kdl is a bad idea unless you have been told it's safe
rennj has joined #haiku
pch_ has joined #haiku
<waddlesplash>
AlwaysLivid: it's as I suspected, a duplicate, and totally unrelated to fs activity anyway
<waddlesplash>
a prospective: fix is already in, hrev55701
<AlwaysLivid>
oof.
<AlwaysLivid>
well, that's awkward.
<AlwaysLivid>
sorry about that.
pch has quit [Ping timeout: 480 seconds]
tqh has joined #haiku
atomozero has quit [Read error: Connection reset by peer]
<waddlesplash>
nah, no problem
<waddlesplash>
this stuff is complicated, I don't expect most people to be able to make any sense of KDL outputs
jesse_blue has quit [Remote host closed the connection]
<AlwaysLivid>
i should seriously consider starting to learn how to read them and try to figure them out, i've been here for a second (not a hot second, but still)
<AlwaysLivid>
eh, i'll manage that with time
KapiX has joined #haiku
KapiX has quit []
<Not-5726>
[haiku/haikuwebkit] pulkomandy pushed 68 commits to webkit2 [+32/-7/±304] https://git.io/JDvV2
<Not-5726>
[haiku/haikuwebkit] RAJAGOPALAN-GANGADHARAN f1f63db - Enable building WebKit2 (non-legacy)
<kallisti5[m]>
something's wrong somewhere in the deep dark collection of rs6000 stuff
<kallisti5[m]>
../src/system/boot/loader/kernel_args.cpp:33:17: error: taking address of packed member of 'kernel_args' may result in an unaligned pointer value [-Werror=address-of-packed-member]
<waddlesplash>
kallisti5[m]: you must pull
<kallisti5[m]>
wait.. didn't you fix those?
<kallisti5[m]>
ah ok
<waddlesplash>
yeah, I did
<waddlesplash>
added -Wno-error= to ArchitectureRules
<kallisti5[m]>
lol.. i guess that works
<kallisti5[m]>
../src/system/boot/platform/efi/dtb.cpp: In function 'void HandleFdt(const void*, int, uint32, uint32)':
<kallisti5[m]>
../src/system/boot/platform/efi/dtb.cpp:488:9: error: 'intc_info' was not declared in this scope
<kallisti5[m]>
hm.. mkay
<waddlesplash>
that'd be a new one, I didn't test building non-x86