<bslsk05>
neugierig.org: Tech Notes: The Success and Failure of Ninja
goliath has joined #osdev
sortiecat has quit [Quit: Leaving]
jcea has quit [Ping timeout: 276 seconds]
errno has joined #osdev
<heat>
oooh ninja finally got GNU make jobserver support
<heat>
fuck yes
<mjg>
lol
Test_User has joined #osdev
xenos1984 has quit [Read error: Connection reset by peer]
\Test_User has quit [Ping timeout: 252 seconds]
xenos1984 has joined #osdev
Test_User is now known as \Test_User
<nikolar>
lol
<nikolar>
google googled
McDonaldsWiFi has joined #osdev
PublicWiFi has quit [Ping timeout: 252 seconds]
<Mutabah>
Random note: yay! I got my OS booting on real hardware... except for the broken USB driver meaning no input devices :(
errno has quit [Ping timeout: 272 seconds]
<n00by>
sweet
Shaddox404 has joined #osdev
<Mutabah>
And maybe I've found the error by reading the linux source... something to check tomorrow morning
Shaddox404 has quit [Quit: (╯°□°）╯︵ ┻━┻]
karenw has joined #osdev
karenw has quit [Quit: Deep into that darkness peering...]
<heat>
i've been implementing scheduler affinity lately
<heat>
migrating a live thread is actually fugly and awful
<heat>
and needs some extra infrastructure i still don't quite have
<k4m1>
Not exactly the right place to ask since fw/'bios' dev, not osdev, but PIC config matches to data-sheets & osdev wiki, software-interrupts & cpu faults trigger my int-handlers, but never getting irq's from eg PIT, qemu reports irq 0 count rising, IMR=F8, IRR=07, ISR=0
<k4m1>
any ideas would be welcome
<k4m1>
right, above on qemu-system-x86_64
<k4m1>
interrupts are enabled (Flag register 0x0..202 )
<k4m1>
feel like am overlooking something really simple, but just can't figure out what exactly have I got wrong ^^`
<k4m1>
on post.c, at line 100, doing pic init, and a few lines later pit init with what i think should be binary-mode square-wave generator at channel 0
<k4m1>
or did I miss something to poke that to trigger irq's?
<heat>
I mean, I don't immediately see what's goign on
<heat>
the mask looks ok, the IRR looks okay, the ISR does not
<heat>
I think
<heat>
it's been a while since i've looked at this sort of stuff out of QEMU
<k4m1>
Right, for PIT, mode-byte is 0x36 ( 0011 0110 )
<k4m1>
ISR looks odd to me too but I can't quite figure out why
<k4m1>
and yea, it's been a few years since did osdev & poked at pic ^^`
<k4m1>
aanyway, thanks for taking a look -- if any more ideas come by, pls ping
<bslsk05>
news.ycombinator.com: Cosmoe: BeOS Class Library on Top of Wayland | Hacker News
* kof673
sees the tree of portability glow silver for a moment, then it returns to normal </nethack>
argantantua has quit [Read error: Connection reset by peer]
goliath has joined #osdev
<pog>
ISR would only be 1 if the interrupt fired and wasn't ACK'd
<pog>
or corresponding to the vector i mean
<pog>
i think i can't remember
<mjg>
spent 2 hours watching fail compilation on youtube
<mjg>
or how i managed to NOT work over the weekend
<heat>
why would you ever want to work on the weekend
<mjg>
DEADLINE
<mjg>
BACKLOG
<mjg>
that aside there are cases where people have a weekend shift to cover production
<mjg>
[not the case here, but just sayin'!]
<heat>
yeah i know
<heat>
it was pretty common at cloudflare to have a guy on the team on-call
<heat>
in case shit fans the hit over the weekend
<heat>
or during the night
<mjg>
so the support dept at rh had rotating shifts for the weekend
<mjg>
where you would be on call if frontline needed help
<mjg>
one day they offered to move regular working days to the weekend for select individuals
<mjg>
there were no takers, go figure
<mjg>
:X
<heat>
:)
<mjg>
ok i'm looking at funny doggos compilation
<mjg>
but so far i'm not laughin'
<mjg>
probably because i don't like dogs
<heat>
mofer what's your opinion on linux scheduler code
<heat>
i find it super hard to follow
<heat>
the locks and the memory barriers
<heat>
it would be so much easier if we replaced it all with, idk, tasklist_lock
<mjg>
i have no opinion
<mjg>
or rather, no basis to have one
<mjg>
but if i had to guess, SOLARIS did it better
<heat>
certainly
goliath has quit [Quit: SIGSEGV]
MelMalik is now known as Reinhilde
karenw has joined #osdev
m3a has quit [Read error: Connection reset by peer]
GeDaMo has quit [Quit: 0wt 0f v0w3ls.]
goliath has joined #osdev
<sortie>
today in sortie can debug anything: i'm losing hair debugging why mediawiki doen't work with crashes taking me through multiple labyrinthine codebases also including php and sqlite
<sortie>
oh for extra fun it crashes after outputting half of the first line of the strack trace
<CompanionCube>
it's in the docs: 'The hard upper bound is automatically set to zero on certain platforms (ex: OpenBSD) where memory mapped I/O does not work due to the lack of a unified buffer cache.'
netbsduser has joined #osdev
<heat>
please don't mention it again or i might have a LMAO attack again
<heat>
and i've LMAOed enough
<sortie>
Looks like I can avoid MAP_SHARED if I disable Write-Ahead Logging in sqlite3
<heat>
add yourself to that LMAO ifdef sortie
<sortie>
heat, you may be laughing, but one of us almost has an OS that runs mediawiki, and the other one is you :P
* heat
laughs in page cache
<zid>
he was laughing at openbsd though?
<sortie>
heat, funny thing about that ifdef is that it appears to be non-functional, it doesn't actually seem to disable memory mapping, fully for real at least
<sortie>
In fact SQLITE_MAX_MMAP_SIZE defaults to 0 on unknown operating systems, so it was already 0 for me