foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
edr has quit [Quit: Leaving]
MiningMarsh has quit [Ping timeout: 264 seconds]
netbsduser has quit [Ping timeout: 252 seconds]
foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
Arthuria has joined #osdev
bradd has joined #osdev
MiningMarsh has joined #osdev
foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
navi has quit [Ping timeout: 268 seconds]
divine has quit [Ping timeout: 256 seconds]
foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
pog has quit [Ping timeout: 240 seconds]
xenos1984 has quit [Read error: Connection reset by peer]
foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
xenos1984 has joined #osdev
pog has joined #osdev
zhiayang has quit [Quit: oof.]
zhiayang has joined #osdev
pog has quit [Ping timeout: 255 seconds]
foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
Matt|home has joined #osdev
foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
Arthuria has quit [Ping timeout: 268 seconds]
rustyy has quit [Quit: leaving]
rustyy has joined #osdev
snappy has joined #osdev
<snappy>
does anyone know how sparse files are implemented in a filesystem? Does there need to be additional metadata that either tracks the allocated blocks or the holes when servicing a read or a write
<Mutabah>
There's already metadata tracking what blocks are used contain the file's data
<Mutabah>
That metadata can have a marker that indicates "no allocated block" instead of "use this block"
<Mutabah>
I'm not 100% sure how the ext2/3/4 family does it, but I suspect sparse regions are represented with a block number of 0...
<Mutabah>
I _do_ know how NTFS does it, and it's with an extent with a physical length of zero (NTFS supports compression, so extents have both a logical length and a physical length)
<geist>
Mutabah: that's precisely how it does it
<geist>
block 0 is 'no block'
<geist>
also i think the main reason block numbers start with 1
<geist>
and/or 0 is otherwise not available
<kazinsal>
NTFS is a wonderfully clever filesystem
<geist>
i dunno how ext4 does it, but i'm gonna guess it's just a regular extent run with a flag saying 'not allocated'
<geist>
kazinsal: but fun! i wrote a set of routines to parse a FILES-11 file system the other day
<kazinsal>
it's fairly well documented in Windows Internals Vol 2 but unfortunately I left my copy at the office so I can't cross-reference it
<geist>
and as i knew, it's pretty much very much NTFS
<kazinsal>
it lives on my desk as a warning to the co-op students to not turn my office into a laptop-flashing station
<bslsk05>
github.com: files11/file.cpp at main · travisg/files11 · GitHub
<Mutabah>
Implementing compressed file support was fun
<geist>
fairly compressed
<kazinsal>
lots of filesystems are fun to fiddle with
<kazinsal>
dunno what I'm going to do for storage when I eventually bother my ass into getting back into this hobby
<geist>
the file records in FILES-11 aren't as free form as they are in NTFS, since the file record really only has 2 or 3 areas, but the start offset of those areas is recorded in the file heeader
<bslsk05>
github.com: files11/ods2.h at main · travisg/files11 · GitHub
spare has joined #osdev
<geist>
side note re ext4 (and probably others), extents have a couple levels of flags too
<geist>
A: this is not present (sparse) and B: this is present but unitialized
<geist>
that lets them allocate a large chunk of a file and not have to zero it out
<geist>
if the B flag is set it just returns zeros on read
<geist>
you can see this with `filefrag -v file`
<geist>
if you just fallocate a large file and dont write to it it'll probably show up as a bunch of unitialized extents
<geist>
i dont know how its encoded in ext4, but i wouldnt be surprised if it's not literally just a flag on the extent structure
<geist>
or a type field or something
<geist>
NTFS (and FILES-11) had a similar thing but simpler: there's a LAST_WRITTEN_TO field in the FILE record that is an offset into the file at which it hasnt' been written to
<geist>
so a new file can be allocated and the last written to field can be set to 0
<geist>
but if you seek to the end of the file and write it then sets it to that offset and has to zero out the whole file
pebble has joined #osdev
aejsmith has quit [Ping timeout: 256 seconds]
j00ru has quit [Remote host closed the connection]
j00ru has joined #osdev
aejsmith has joined #osdev
ZombieChicken has joined #osdev
ZombieChicken has quit [Client Quit]
ZombieChicken has joined #osdev
j00ru has quit [Ping timeout: 268 seconds]
basil has quit [Ping timeout: 272 seconds]
j00ru has joined #osdev
<nikolar>
geist, if you're here, my bsd2.11 is apparently at patch 195
<nikolar>
Since you asked
<geist>
gotcha
<geist>
you gotta long way to go!
<geist>
but honestly the patches are kinda fun to apply
<nikolar>
Yeah I installed it at that patch level
<nikolar>
I don't know, might give it a go to run all other patches
<nikolar>
Where do I even find then heh
<geist>
i think i started at 451 and manually patched to 481
<geist>
on a real machine so it took like maybe 15-20 hours total
<nikolar>
Heh only 30
<nikolar>
Imagine all 400
<geist>
since building the kernel takes like 30 minutes or so
<nikolar>
Well I hope you wouldn't have to rebuild the whole thing if the patch only touches a few files
<geist>
most of the patches have specific instructions about what is effected and what you need to rebuild
<geist>
so usually it's build a handful of programs, or if it's a kernel thing, rebuild the kernel
<geist>
and yeah if you leave the old .o files around it'll do an incremental
<geist>
though if it touches a header or whatnot it might rebuild the whole thing
<geist>
but your simh will run rings around real hardware. like 50x
<geist>
unless you aggressively throttle it
<nikolar>
Heh I usually run it throttled when it's running the irc client, but when I expect compiling or frequent restarts, I just let it rip
ZombieChicken has quit [Quit: WeeChat 4.3.2]
gbowne1 has quit [Read error: Connection reset by peer]
<geist>
yeah and even then the throttling is weird,since it's instructions per second, and a lot of the pdp11 instructios are extremely variable cycle wise
<geist>
so it's highly load dependent
<nikolar>
Heh yeah guess that makes sense
<nikolar>
Probably would've been more expensive to track actual cycles
spare has quit [Ping timeout: 272 seconds]
GeDaMo has joined #osdev
pog has joined #osdev
pog has quit [Client Quit]
spare has joined #osdev
CryptoDavid has joined #osdev
pebble has quit [Read error: Connection reset by peer]
Left_Turn has joined #osdev
<nikolapdp>
geist: one funny thing, ps doesn't work on my pdp because i don't have swap heh
<nikolapdp>
it just dies saying /dev/swap doesn't exist
<nikolapdp>
and if i give it an empty file it prints out one process and then it dies
<nikolapdp>
*empty file for swap
<nikolapdp>
also one question, is there a canonical page for bsd patches
baraq has joined #osdev
aejsmith has quit [Ping timeout: 255 seconds]
aejsmith has joined #osdev
spare has quit [Remote host closed the connection]
spare has joined #osdev
nikolapdp has quit [Remote host closed the connection]
vdamewood has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
nikolapdp has joined #osdev
<nikolapdp>
ILLUMOS
Tetsuo has joined #osdev
navi has joined #osdev
Left_Turn has quit [Ping timeout: 268 seconds]
Tetsuo has quit [Quit: WeeChat 4.3.2]
netbsduser has joined #osdev
Left_Turn has joined #osdev
Turn_Left has joined #osdev
Left_Turn has quit [Ping timeout: 255 seconds]
spare has quit [Remote host closed the connection]
nikolapdp has quit [Remote host closed the connection]
<mjg>
did someone say PIECE OF SHIT KERNEL
<nikolar>
SOLARIS
<mjg>
did someone say SCALABLE KERNEL
edr has joined #osdev
foudfou_ has joined #osdev
foudfou has quit [Ping timeout: 260 seconds]
q3lont has joined #osdev
foudfou_ has quit [Remote host closed the connection]
foudfou has joined #osdev
<dostoyevsky2>
mjg: dragonflybsd is said to have a scalable kernel
basil has joined #osdev
<mjg>
did someone say DEAD BSD
<mjg>
albeit DEAD + BSD is kind of a pleonasm innit
aejsmith has quit [Ping timeout: 240 seconds]
<dostoyevsky2>
nothing scales better than death
foudfou has quit [Remote host closed the connection]
goliath has joined #osdev
foudfou has joined #osdev
hussein is now known as nur
aejsmith has joined #osdev
bitoff has quit [Remote host closed the connection]
nikolapdp has joined #osdev
<nikolapdp>
BSD2.11 KERNEL
<mjg>
did someone say GENIUS
<nikolapdp>
indeed
<mjg>
indeed? i don't know that system
<nikolapdp>
NT
<mjg>
relax man you may be triggering some people now
<mjg>
you like to steer shit up don't you
<nikolapdp>
me, never
<mjg>
my mistake
<nikolapdp>
you are forgiven
<mjg>
> WARNING: Could not find kernel headers in or .
<mjg>
great error messages
<nikolapdp>
what's that
<mjg>
with all the shitting i'm doing on SUN ENGINEERING ETHOS
<mjg>
i have to give them credit where credit is due
<mjg>
the user-visible side of tooling was very well polished
<mjg>
the things you needed were reasonably handy to execute, error messages were sensible and so on
<mjg>
for example with dtrace
<mjg>
in contrast i'm messing with bpftrace at the moment and i keep finding lul bugs
<mjg>
and the overall user experience is absolute crap
<bslsk05>
github.com: Fix func builtin in kretprobes and uretprobes by ajor · Pull Request #2645 · bpftrace/bpftrace · GitHub
<mjg>
debian does not havea backport, so i compiled it from scratch
<mjg>
this is where it failed to startup with the above helpful message
<nikolapdp>
kek
<mjg>
even ignoring that aspect it takes a stupid amoutn of time to instrument probes
<mjg>
and to bugger off after
<mjg>
while not a good idea per se, with dtrace even on freebsd you can safely instrument almost the entire fucking kernel
<mjg>
meanwhile this bugger almost tapped out at 2k probes
<mjg>
also it somehow requires more than 1K fd limit
<nikolapdp>
BPF
<mjg>
if you stick to the default of 1024 you get a flood of "lmao too many open files"
<nikolapdp>
eBPF rather
<mjg>
at the moment this thing is a tech preview
<mjg>
dtrace reached more maturity in the shorter timeframe
<nikolapdp>
who uses ebpf i wonder
<mjg>
well i'm trying
<mjg>
some people claim they do
<mjg>
so far anything but super trivial usage fails
<mjg>
the best i got working so far is simple aggregations of a param when hitting a probe
<mjg>
anything fancy and things go to shit, example above
<mjg>
i found that oracle is keeping their dtrace for linux port almost up to date
<mjg>
caught up to 6.7 kernel
<mjg>
i'm this close to trying it out just to get the instrumenation i need
<nikolapdp>
huh neat
<mjg>
seriously, solaris == shite perf, but high quality experience for their bespoke tooling
<mjg>
there are portugal fucks who don't like zfs, but people like that aside people mostly do like it
<mjg>
zfs has its warts for sure, but was way ahead of the competition back in the day
<nikolapdp>
it still is in a lot of ways
<nikolapdp>
i use it
<mjg>
ye zfs is pretty nice
<mjg>
from end user/sysadmin pov
<mjg>
internals are shite from scalability pov
<mjg>
i know, i patched some of them
<nikolapdp>
eh maybe, probably doesn't matter on hdds
node1 has joined #osdev
<nikolapdp>
not sure how widely used it is with ssds
<netbsduser>
if solaris is slow now it is mainly because it's been dead for 1 and a half decades
<mjg>
it very much does matter if you have your working set cached
<nikolapdp>
that's also not wrong
<netbsduser>
in its day it was called "the queen of unixes"
<mjg>
and are operating past laptop scale
<mjg>
netbsduser: look mate, it had its charm and some decent user-facing tech, but internals overall are classic dogshit unix
<mjg>
they are suffering from abundance of wtf
<netbsduser>
we were all dogshit unixes once
<mjg>
my unix clone, which i never managed to get past working toy interrupt handling, was never dogshit
<netbsduser>
i will say it is a lot of the user-facing stuff that really is the jewel in the crown
<netbsduser>
dtrace is well regarded with good reason
<mjg>
i see why from user/admin pov not knowing what's under the hood one would be impressed
<nikolapdp>
also smf is regarded as the best init to have ever existed
<mjg>
ignorance is bliss
<nikolapdp>
by some
<netbsduser>
nikolapdp: yes, i think i have some credentials to speak on this topic (having ported systemd to the BSDs) and i do agree
<mjg>
look man. i use vim (now neovim)
<nikolapdp>
heh i think we've had this conversation netbsduser
<mjg>
i think it's a great editor
<mjg>
i also know the internals are fucking disgusting
<nikolapdp>
just switch back to vim :P
<mjg>
so i pretend it's not a thing
<netbsduser>
you can see the vaunted sun engineering ethos in the smf codebase, that was my first exposure to solaris code and that might in large part be the basis of why i will always be a little bit irrationally fond of it
<mjg>
let me tell you a funny story mate
<mjg>
when i was young and handsome, as opposed to just handsome, i was just starting messing with multicore
<mjg>
hearing a lot about solaris i was dlighted to dig into the source code of their mutex impleentation
<mjg>
at the time i found it very elegant, concise and overall great quality
<mjg>
some time later my multicore-clue settled in and when revisiting the code i found that it they instead demonstrated either lack of understandign what they are doing
<mjg>
or gross negligence
<mjg>
either way code being CRAP
foudfou has quit [Remote host closed the connection]
<mjg>
i wrote a detailed critique somewhere
foudfou has joined #osdev
<nikolapdp>
fight fight fight
<netbsduser>
i can't say i've ever looked at the actual implementation, what always wound me up was that they have a mutex struct and it's either a sleeping lock (with adaptive spinning) or a spinlock (with nothing but spinning)
<mjg>
that is a problem, but it is merely a side note to the actual criticism
<mjg>
somehow i can't find my write up
<netbsduser>
it's my own little aesthetic irritation with it, i don't understand why they did it
<mjg>
bottom line though is that they do trivially avoidable lock accesses, some of which are super harmful
<mjg>
the entire thing not passing a sniff test for a serious locking primitive
op has joined #osdev
Arthuria has quit [Ping timeout: 264 seconds]
goliath has quit [Quit: SIGSEGV]
q3lont has quit [Quit: Leaving]
<nikolapdp>
KERNAL
CryptoDavid has quit [Quit: Connection closed for inactivity]
FreeFull has joined #osdev
foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
<snappy>
i'm not sure how to phrase this question, but does showing a keen interest/exploration in pedagogical OSes such as xv6, xinu, pintos, the one that Drew DeVault came out with recently look undesirable if you were applying for say a linux kernel programming job (i know this could mean a lot of things given how vast the linux kernel is)?
<bslsk05>
chalaux.org: Chapter 5. Aristotle against Plato. The power of money.
<kof673>
point: anything can mean anything depending on what type of logic is permitted
<FreeFull>
snappy: Guessing you're applying to that job at Canonical? They explicitly ask for any osdev experience, so probably
<FreeFull>
Well, it'd be best if you wrote something osdev-related yourself
<nortti>
surely that would look desirable, not *un*desirable, if they're asking for it?
<snappy>
FreeFull: spot on, yeah i don't have kernel experience in a professional capacity, but i really like looking at these smaller OSes as they're easier to experiment with and there's a lot of good pedagogical resources
node1 has quit [Quit: Client closed]
goliath has joined #osdev
simpl_e has quit [Ping timeout: 255 seconds]
<nikolapdp>
VMUNIX
aejsmith has quit [Ping timeout: 256 seconds]
nyah has joined #osdev
k_hachig has joined #osdev
theruran has joined #osdev
foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
aejsmith has joined #osdev
op has quit [Remote host closed the connection]
foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
heat has joined #osdev
<heat>
good morning operating systems
dalme has joined #osdev
<heat>
i no longer feel like complete shit
<nikolapdp>
what did you do
<heat>
i was sick with the flu for the past 4 or 5 days
<heat>
it was awful
<nikolapdp>
ah yeah
<mjg>
heat: want something to cheer you up
<heat>
today is the first day i don't have a fever, which is gr8 m8
<bslsk05>
<corbet> On the radar: using a large-language model to insert thousands of automatically generated "security checks" into the OpenBSD kernel:https://marc.info/?l=openbsd-tech&m=171810103406609&w=2I'm sure that will be received well...
<heat>
i've seen that
<heat>
very funny stuff
<mjg>
dude claims patched files are "hardcored"
<mjg>
i'm totally going to use it
<heat>
if you do the same checks more often it's totally safer
<mjg>
these are not the same checks
<mjg>
it's stuff like crap->lolvar -= lolparam;
<mjg>
replaced with if (crap->lolvar < lolparam) crap->lolvar = 0; ....
<mjg>
overall fucking brilliant
<nikolapdp>
lol
<nikolapdp>
what if you want it to go negative
<nikolapdp>
or error on negative
<mjg>
llm knows better you fuck
<nikolapdp>
of course
<nikolapdp>
how could i forget
foudfou has quit [Remote host closed the connection]
<heat>
geist: yeah i think it was the flu, my mom is bedridden now but my sister got the flu shot a few months back and still isn't sick
<heat>
could also be covid, the doctor didn't really care, she just knew it was some virus
<heat>
but it didn't feel covid-ish to me
<mcrod>
nikolapdp this is nuts though
<mcrod>
this means they don't test their shit
<mcrod>
assuming I didn't do something stupid
<nikolapdp>
yeah no clue
<mcrod>
note that I've done exactly this on a Cortex-m4 and no surprise it works
<mcrod>
so what the fuck am I supposed to do?
<mcrod>
i complain a lot for only one reason
<mcrod>
this stuff shouldn't happen from a company like ARM.
<heat>
you're complaining about linaro
<heat>
not ARM
<mcrod>
ok linaro
<geist>
are you using the right libgcc, etc?
<mcrod>
i'm using exactly what they provide
<geist>
or more specifically do they have a multilib for cortex-m0?
<geist>
what does your toolchains multilib map look like?
<mcrod>
one sec
<geist>
i forget the switch but it's something like --print-multilib
<mcrod>
i need to download it on my home machine
<mcrod>
sec
<kof673>
is the offending code assembly? if not, could just be a bad flag leaked in somewhere (on their side -- i.e. packaging) this is rhetorical, just maybe prior version worked
<kof673>
which does not excuse "not testing" but ....maybe it did work once...
<geist>
well i guess my point is the multilib map needs to cover armv6m somehow
<GeDaMo>
Do you know what the mrc2 instruction is being used for?
<geist>
if it doesn't then they just didn't provide all the variants to run on cortex-m0 and m0+ which are v6m
<geist>
vs v7m and v8m
<mcrod>
GeDaMo part of the libgcc _fini destructor init bullshit
<GeDaMo>
"Many of the supported CPUs implement optional architectural extensions. Where this is so the architectural extensions are normally enabled by default."
<bslsk05>
gcc.gnu.org: ARM Options (Using the GNU Compiler Collection (GCC))
<mcrod>
guess what people
<GeDaMo>
Maybe try something like -mcpu=cortex-m0plus+nofp
<mcrod>
I will be very angry if +nofp is actually what I need
<mcrod>
because I'm 99.99% sure FPUs don't exist on Cortex-M0
<mcrod>
so that should be implied
<heat>
have you tried passing -march explicitly
<geist>
that's correct
<geist>
yeah you should pass arch such that it selects the right libgcc
<mcrod>
but -mcpu is supposed to
<heat>
also are you linking correctly? with gcc
<mcrod>
do not mistake me, i'll try anything
<geist>
mcrod: yes but thats what the multilib map shows you
<mcrod>
and right now I can't test it anyway since I bricked the dev board :)
<geist>
it shows you how it tests for the right libgcc
<mcrod>
my life is a struggle
<geist>
so in this case it's not matching against mcpu. usually doesn't
<mcrod>
so you're saying -mcpu=cortex-m0plus+npfp might actually be the right thing?
<nikolapdp>
i think mcpu is depracated?
<nikolapdp>
or which one was it
<mcrod>
not for ARM as far as I know
<geist>
-march=armv6s-m it looks like
<geist>
note this is all pretty standard, my toolchains have exactly the same multi lib map. it's the default multilib map in gcc
<geist>
-mcpu may select it, but you gotta experiment to see what makes it hapapy
<mcrod>
that's nuts
<geist>
having that many multilib variants is nuts, but that's because arm32 is nuts
<mcrod>
note: i believe you
<mcrod>
but that's actually really insane
* geist
shrugs
<mcrod>
*if* it works, and again I can't test it
<mcrod>
right now, anyway
<geist>
i dunno it's just how it's always been
<geist>
been that way for 20 something years
<mcrod>
well I've never had to really sink my teeth into this for ARM until today
<mcrod>
:p
<mcrod>
i can see in the map file what it used at least..
<mcrod>
it did use thumb/v6-m/nofp/libg.a, so that sounds right so far
<mcrod>
so let's see..
<nikolar>
you could see if the problematic instructions is there
<mcrod>
yes
<mcrod>
that is indeed my plan
<mcrod>
but I'm also going to use -frecord-gcc-switches
<mcrod>
ok -march=armv6s-m was passed
joe9 has joined #osdev
<mcrod>
so I'm still in the game
<nikolapdp>
nice
<mcrod>
hm...
<mcrod>
where the fuck is _fini
<heat>
_fini is in crtend.o
<heat>
well, _fini is crtn.o, .fini (the interesting stuff) is in crtend.o
<mcrod>
there's nothing fucking in there
<mcrod>
well
<mcrod>
now I'm very fucking confused
<mcrod>
I don't see that instruction anywhere..
<mcrod>
but gdb said it was there
<mcrod>
god
<mcrod>
when I run my .elf through objdump -S, I see the first two instructions are correect
<mcrod>
the 3rd one is a pop, not an mrc2
<heat>
you're loading it wrong?
<mcrod>
i don't see how
<mcrod>
if you're asking how I'm flashing it, J-Link is smart enough to figure out what to do based on the target board and the contents of the ELF header
<mcrod>
this is what kills me too
<mcrod>
that's the last hurdle before it calls main
<mcrod>
fuck
<heat>
i bet your j-link thing is borked
<mcrod>
we use it regularly every day
<mcrod>
i doubt it
<mcrod>
it's also not even a "thing"
<mcrod>
it's like a staple in embedded development
<mcrod>
hopefully it's something idiotic that isn't my fault
<heat>
listen, your flash contents allegedly aren't matching your final executable
<heat>
so check that
<mcrod>
i will tomorrow, can't fuck with it today
<mcrod>
i'm sorry, I shouldn't be this frustrated
josuedhg has joined #osdev
vikn has left #osdev [#osdev]
bauen1 has quit [Ping timeout: 240 seconds]
bauen1 has joined #osdev
netbsduser has quit [Read error: Connection reset by peer]
heat has quit [Quit: Client closed]
spare has joined #osdev
qubasa has quit [Remote host closed the connection]
sjs has quit [Ping timeout: 260 seconds]
sjs has joined #osdev
gog has joined #osdev
foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
k0valski18891621 has quit [Quit: Peace out !]
gildasio has quit [Ping timeout: 260 seconds]
aejsmith has quit [Ping timeout: 268 seconds]
<Ermine>
Oh, I've got new features with bios upgrade
<GeDaMo>
Are they things you want? :P
<Ermine>
No, unfortunately. But "power efficient turbo mode" sounds cool nonetheless
divine has joined #osdev
aejsmith has joined #osdev
gildasio has joined #osdev
<Bitweasil>
Ermine, ODroid H4?
<Ermine>
Bitweasil: no
<Bitweasil>
My guess, rather speculative, is that it will recognize that the CPU is waiting on data enough that it may as well clock down and save some power, but I'm really not sure.
<Bitweasil>
Turbo burns a *lot* of power for a fairly marginal gain in performance, and I end up with it disabled on a lot of my hardware.
foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
<Ermine>
Maybe you're right, but that's above my paygrade
m3a has quit [Ping timeout: 264 seconds]
foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
m3a has joined #osdev
foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
spare has quit [Remote host closed the connection]
op has joined #osdev
dalme has quit [Read error: Connection reset by peer]
joe9 has quit [Quit: leaving]
GeDaMo has quit [Quit: 0wt 0f v0w3ls.]
foudfou has quit [Remote host closed the connection]
foudfou has joined #osdev
gbowne1 has joined #osdev
gbowne1 has quit [Remote host closed the connection]