Search logs:

channel logs for 2004 - 2010 are archived at ·· can't be searched

#osdev2 = #osdev @ Libera from 23may2021 to present

#osdev @ OPN/FreeNode from 3apr2001 to 23may2021

all other channels are on OPN/FreeNode from 2004 to present

Wednesday, 6 September 2023

00:01:00 <zid> camels suck, how about a dragon
00:02:00 <x8dcc> camels don't suck
00:02:00 <vdamewood> zid: I have a book with a dragon on it. A purple dragon.
00:04:00 <zid> does it have malignalitaloptereosis?
00:12:00 <jimbzy> Our bible would have a capybara on the cover.
00:13:00 <x8dcc> Hell no
00:13:00 <zid> yea, nah
00:13:00 <zid> capy are friend shaped
00:13:00 <zid> osdev is not friend shaped
00:13:00 <zid> it's self-loathing shaped
00:13:00 <x8dcc> my back hurts
00:13:00 <x8dcc> my eyes too
00:14:00 <jimbzy>
00:14:00 <bslsk05> ​ <no title>
00:15:00 <x8dcc> jesus it's ugly
01:50:00 <geist> yay osdev people are weird people with a community
01:52:00 <kazinsal> mmmmyep
02:16:00 <zid> I think I have malignalitaloptereosis
12:30:00 <x8dcc> I was literally going outside and then I remembered something. About the debug exception issue I was talking about yesterday, Since IRETD restores EFLAGS, couldn't I just unset the trap flag bit when I enter the ISR and then IRETD would restore the old ones?
12:30:00 <x8dcc> that way I think it would not debug itself
12:30:00 <x8dcc> I still need to leave so I will check it later
13:15:00 <Ermine> gog: may I pet you
13:22:00 <gog> yes please
13:22:00 * Ermine pets gog
13:23:00 * gog prr prr
13:26:00 <ChavGPT> i watched a video about cats and comments say it is lying
13:26:00 <ChavGPT> video claims cats check up on you in the morning
13:26:00 <ChavGPT> and it is totally not about food
13:26:00 * ChavGPT throws gogs a snack
13:26:00 <ChavGPT> what is your take gogs
13:28:00 <gog> they're checking to see if you're dead because if you are then you're the food
13:29:00 * ChavGPT is happy to not own a gogs
13:35:00 <gog> speaking of cats i pet an orange kitty on the way to work this morning
13:35:00 <gog> v friendly 13/10 would pet again
13:36:00 <ChavGPT> maybe cultures eating cats have it right
13:36:00 <ChavGPT> SAME WITH DOGS
13:36:00 <sham1> No >:(
13:37:00 <ChavGPT> ey sham1 do you ever eat beef
13:37:00 <ChavGPT> cause lemme tell you, cows can be very friendly!
13:37:00 <sham1> Yes they can be
13:37:00 <sham1> I've abstracted away the murder
13:39:00 <Hammdist> is there a syntax to read ttbr0_el1 from gdb connected to qemu (on aarch64)?
13:46:00 <mcrod> hi
13:46:00 <mcrod> gog may I pet you
13:46:00 <mcrod> zid: zid: hey zid! almost got audio working correctly except, pops
13:46:00 <mcrod> but only during one aspect of generation
13:48:00 <gog> yes
13:48:00 * mcrod pets gog
13:48:00 * gog prr
13:50:00 <x8dcc> suspicious(tm)
13:51:00 <mcrod> ok you caught us
13:51:00 <mcrod> gog and I hold hands sometimes
13:51:00 <x8dcc> kinda weird thing to say on the internet but I ain't judging
13:54:00 <mcrod> uh
13:54:00 <mcrod> rest assured that much weirder things go on with the internet
13:55:00 <jimbzy> Interesting... My laptop shutdown during the middle of the night.
13:55:00 <mcrod> that was me, sorry.
13:55:00 <jimbzy> Ahhh. I see the problem. I forgot to plug it in. XD
13:56:00 <jimbzy> A standard ID10T fault.
13:57:00 <mcrod> god I hate audio
14:02:00 <gog> h*ld h*nds o//o
14:04:00 <ChavGPT> rust assured
14:04:00 <ChavGPT> so i just watched another cat video
14:04:00 <ChavGPT> one of videos earlier today: why can't sleep with you
14:04:00 <ChavGPT> ez -- they protect you bro
14:04:00 <gog> protecc
14:04:00 <ChavGPT> now the one i got: they feel safe bro
14:05:00 <ChavGPT> basically pick whatever narrative makes you feel better and roll with it
14:05:00 <gog> basically
14:05:00 <gog> there's a myth that cats don't meow to communicate with one another
14:05:00 <gog> but i've seen videos of a cat with a collar cam and they meow at one another all the time
14:05:00 <gog> it's certainly not as much but it does happen
14:06:00 <mcrod> today my cat knocked over a glass bulb.
14:07:00 <ChavGPT> i'm waiting for someone offended at the idea their cat it like other cats
14:08:00 <ChavGPT> s/at/by/
14:12:00 <zid> my cat goes up to other cats and says in a very deep voice "Oi, fucker"
14:14:00 <mcrod> yay I fixed it
14:14:00 <mcrod> thanks zid
14:14:00 <mcrod> <- needs to be patted on the head but it works
14:14:00 <bslsk05> ​ samething/src/core/src/private/core.c at main · mcroddev/samething · GitHub
14:53:00 <Hammdist> if I update the memory contents of page tables, is `tlbi vmalle1is ; dsb sy ; isb` sufficient to make the new mapping effective?
15:03:00 <Hammdist> linux uses a different sequence but plugging that in does not fix my current issue
15:03:00 <bslsk05> ​ tlbflush.h - arch/arm64/include/asm/tlbflush.h - Linux source code (v6.5.1) - Bootlin
15:04:00 <Hammdist> I am running in QEMU and while I can "walk" the page tables manually using xp command and everything looks fine, the emulator doesn't like something about it
17:09:00 <zid> gog: Found your mastadon
17:10:00 <bslsk05> ​Note by Sophie 🎲🧩🏳️‍⚧️: "Just realised my prescriptions rn consist of: ␤ - Antibiotics 🚫🦠 ␤ - Anti-boy-otics 🚫🧔‍♂️ ␤ - Anti-boo-otics 🚫😭"
17:11:00 <x8dcc> what am I looking at
17:30:00 <gog> i'm not on antibiotics rn
17:31:00 <x8dcc> we need Jesus more than ever
17:32:00 <gog> nah praise satan
17:32:00 <Ermine> ^
17:35:00 <zid> church of satan is pretty cool
17:35:00 <bnchs> imagine having a religion
17:35:00 <bnchs> lol, go areligious
17:36:00 <bnchs> i don't care if there's a god or multiple gods or not
17:38:00 <gog> i jsut want this fucking sales portal to work
17:38:00 <Ermine> x8dcc: please re-read Matthew 7
17:39:00 * Ermine cheers up gog
17:40:00 <gog> thanks
17:40:00 <gog> ¯\_(ツ)_/¯ it works in dev but not production
17:40:00 <gog> not clear why
17:40:00 <zid> Sounds like it works fine then
17:40:00 <zid> production is bad
17:40:00 <gog> true
17:40:00 <x8dcc> Ermine: you are right
17:50:00 <sham1> HAIL SATAN
17:50:00 <zid> sham1: is that a magic spell that makes it rain solid chunks of satanism?
17:51:00 <sham1> Yes
18:01:00 * kof123 gathers the magical raining sulphur, mixes with magical rotting code
18:15:00 <heat> if u close your eyes and say this in front of a mirror i'll pop up right behind you
18:16:00 <heat> mcrod: really like you dumping the whole license onto every file
18:16:00 <heat> really brings out the GNU vibes
18:16:00 <heat> even with the SPDX license stuff
18:25:00 <gog> heat
18:25:00 <gog> heeeeeeat
18:29:00 <heat> gog
18:29:00 <heat> gog
18:29:00 <heat> gooooooooooooooooooog
18:29:00 <gog> gog
18:29:00 <heat> heat
18:29:00 <heat> heat fuck you u suck
18:30:00 <heat> heat: bazinga
18:33:00 <ChavGPT> heat practicing for his lkml appearance
18:36:00 <zid> heat: Can you go away for a couple of hours? I'm sweating
18:38:00 <heat> ChavGPT: shuddup mr pessimal
18:38:00 <heat> go pessimal all over your commit messages
18:39:00 <ChavGPT> my commit messages are THE SHIT
18:39:00 <heat> btw the "kernel test robot noticed a -8.2% improvement of phoronix-test-suite.osbench.LaunchPrograms.us_per_event on" you got today really fucked me up
18:39:00 <heat> i took some time to grok that one
18:40:00 <ChavGPT> L M A O innit
18:40:00 <ChavGPT> personally i'm offended by a phoronix win
18:40:00 <heat> you are now the michael larabel
18:40:00 <heat> michal larabelsz
18:40:00 <ChavGPT> Pedro Krause
18:40:00 <ChavGPT> ever heard of nick krause?
18:41:00 <ChavGPT> seach for the guy on lkml kek/10
18:41:00 <ChavGPT> it could have legitimately be a troll
18:41:00 <ChavGPT> been
18:42:00 <zid> -8% improvement is what I expect from you two
18:43:00 <heat> tfw zid's shell script runs 8% slower because now it uses fancy-ass algorithms and data structures to scale to 50000000 cpus
18:44:00 <zid> ikr
18:44:00 <zid> give rep scasb back
18:45:00 <zid> "sorry, too busy pulling the ripcord on the avx512 engine to run your code"
18:45:00 <zid> "damn thing doesn't wanna start, must be cold, come back when the glow plugs are hot"
18:45:00 <zid> computers are a disgrace
18:51:00 <sham1> Computers are GOATed
18:51:00 <sham1> 🐐
18:53:00 <mcrod> hi
18:53:00 <mcrod> heat: <3
18:53:00 <heat> mcrod is GOATED
18:54:00 <zid> I don't think you should be calling each other welsh like that
18:54:00 <mcrod> anyway, fuck audio and I'm glad I figured it out.
18:58:00 <sham1> We can hear
19:01:00 <gog> computers are goated with the sauce
19:02:00 <heat> computers are rizzed up
19:03:00 <zid> I don't think even you guys know what you're saying at this point
19:08:00 <heat> bold of you to assume i ever did
19:15:00 <geist> interesting writeup on a new sifive core announced:
19:15:00 <bslsk05> ​ Hot Chips 2023: SiFive’s P870 Takes RISC-V Further – Chips and Cheese
19:15:00 <geist> looks like they're getting pretty aggressive now
19:16:00 <nikolar> could be interesting
19:16:00 <nikolar> but they are directly competing with arm so who knows
19:16:00 <geist> yeah that core lookssl ike a direct competitor to a high end arm core
19:17:00 <heat> oh fun tidbit i saw today
19:17:00 <heat> intel is coming up with family 16h
19:17:00 <heat> i don't know what for though
19:17:00 <nikolar> family 16h?
19:17:00 <heat> yeah, cpuid family
19:17:00 <nikolar> ah right
19:17:00 <zid> i7dx
19:17:00 <geist> huh they've been stuck on their previous family forever
19:17:00 <Ermine> heat: how are you?
19:18:00 <geist> i guess that doesn't really collide with AMD's family 16h since it's a different vendor
19:18:00 <zid> maybe we finally get fred
19:18:00 <geist> i think what, intels been using family 0xf forever?
19:18:00 <nikolar> wonder why they feel the need to that right now
19:18:00 <geist> possible theyre switching families because they literaly ran out of space. iirc there's been this one 8 bit space they've been filling in
19:18:00 <heat> Add New Intel Processor family for SMBIOS Type 4
19:18:00 <heat> Hex value - 16h
19:18:00 <heat> Name - Intel® Processor
19:18:00 <heat> in tianocore
19:18:00 <geist> and iirc it's getting fairly full
19:19:00 <nikolar> interesting
19:20:00 <geist> also i forget, itanium used up a family id i think
19:20:00 <geist> the way intel maps cpu implementations to family is a complicated mess. AMD is much more straightforward
19:20:00 <heat> huh now im wondering if this is just some smbios crap
19:20:00 <heat> it says processor family which I assumed matched up with CPUID
19:21:00 <geist> i'll ask around at work, thjough may not be able to say anything based on what the answer is
19:21:00 <heat>
19:21:00 <bslsk05> ​ edk2/MdePkg/Include/IndustryStandard/SmBios.h at master · tianocore/edk2 · GitHub
19:21:00 <geist> ah yeah thats just some enum
19:21:00 <geist> and for whatever reason 16h wasn't used before
19:21:00 <gog> fjölskyldutala
19:22:00 <zid> lga4710-2 inc though
19:22:00 <sham1> gog: bless you
19:22:00 <zid> so they can glue more bad i7s together to sell
19:22:00 <geist> but for similar reasons it looks like, that enum is running out of space assuming it's fixed to 8 bits
19:22:00 <geist> so probably going back and filling in the gaps
19:23:00 <geist> i got a new external backup drive, now i begin the slow process of filling it with random garbage before formatting it with dmcrypt
19:23:00 <geist> going to take....16 hours
19:27:00 <nikolar> kek
19:27:00 <mcrod> i'm nauseous.
19:33:00 <gog> backups?
19:33:00 <gog> pffff
19:49:00 <mcrod> heat:
19:49:00 <bslsk05> ​ samething/src/frontend/sdl/main.c at main · mcroddev/samething · GitHub
19:53:00 <zid> is that our fourcc
19:54:00 <zid> also, that's a fairly tight sample buffer, 23ms?
19:55:00 <mcrod> yeah but so far it's working fine
19:55:00 <mcrod> i should increase it to 16,384
19:55:00 <zid> now do it with prime95 running
19:55:00 <zid> that's too big :P
19:55:00 <mcrod> i'll show you what's too big >:(
19:55:00 <zid> you have no sense of proportion do you
19:55:00 <mcrod> no.
19:56:00 <zid> 16 * 23 = 368ms
19:56:00 <zid> that means click button, third of a second later, audio plays
19:56:00 <zid> 23ms is just *slightly* small
19:56:00 <zid> not 16x too small
20:31:00 <ornx> does something special in x86-64 go at address 0xf? is that real memory? my EFI memory map i get after exiting boot services says there's a block of memory there 12288 bytes long...
20:34:00 <ornx> 3x4096... 12x1024...
20:34:00 <heat> ornx: no
20:35:00 <heat> i'm willing to bet you're reading the memory map wrong
20:35:00 <heat> 0xf sounds very bogus
20:35:00 <heat> pls show code and struct definition
20:37:00 <zid> yea you're either offset by some amount, or you overread, or you read the flags as the address, presumably
20:41:00 <heat> EFI has a fun detail where the map descriptor's size is an out-parameter of GetMemoryMap
20:41:00 <heat> you *CANNOT* do EFI_MEMORY_DESCRIPTOR *desc = ...; for (;; desc++)
20:42:00 <heat> rather for (u8 *ptr = descriptors; /* some condition i cant remember */; ptr += desc_size)
20:42:00 <zid> The multiboot one is described using that weird negative format too
20:42:00 <zid> why's it gotta be so weird in general
20:43:00 <heat> so it's extensible
20:43:00 <ornx> src/amd64/entry.c:689 assertion failed: desc_size == sizeof(EFI_MEMORY_DESCRIPTOR)
20:43:00 <zid> negative indexing does not make something sextensible
20:43:00 <heat> tianocore actually has a funny gotcha there where they increment the descriptor size
20:43:00 <heat> ornx: ^^
20:43:00 <heat> it will not match sizeof(EFI_MEMORY_DESCRIPTOR)
20:43:00 <heat> on purpose
20:44:00 <ornx> ahh
20:44:00 <heat> the clever cunts added a gotcha so everyone used desc_size
20:44:00 <ornx> wait, how do you index the fields inside of it then
20:44:00 <heat> bruv
20:44:00 <heat> for (u8 *ptr = descriptors; /* some condition i cant remember */; ptr += desc_size) {
20:44:00 <heat> EFI_MEMORY_DESCRIPTOR *desc = (void *) ptr; }
20:45:00 <heat> i have no idea if my shit irc client made the void pointer into a smiley
20:45:00 <heat> but if it did, it's pretty funny
20:45:00 <zid> It did not
20:45:00 <heat> aw
20:46:00 <ornx> ah so you can use the struct offsets, just not the struct sizeof
20:46:00 <heat>
20:46:00 <bslsk05> ​ <no title>
20:46:00 <ornx> that's fine then
20:46:00 <heat> yes
20:46:00 <heat> ofc blah blah be careful and don't access beyond the actual size
20:46:00 <zid> I'd delete that irc client immediately
20:46:00 <heat> but i don't think the memory map descriptor has been expanded, ever. might be wrong though
20:46:00 <heat> zid: i'm on since i'm on windows and i don't do nerd programming on windoze
20:50:00 <zid> yea, delete
20:50:00 <zid> or install wsl and irssi, or mirc
20:50:00 <heat> i don't want to have a full linukz running for irssi
20:51:00 <heat> an idle WSL2 system with bash open balloons like half a gig or something, last i checked
20:52:00 <zid> yea mine's using 645MB, but, so what?
20:52:00 <zid> That's 1/64th of my ram
20:52:00 <zid> and windows *loves* swap, so I can probably run a hundred of them np
20:54:00 <heat> yeah but i have 8 gigs
20:54:00 <zid> why
20:54:00 <zid> did the other 24 fall out
20:54:00 <heat> its a shitty laptop from 2018
20:55:00 <zid> Did the other 8 fall out then
20:55:00 <heat> it was (back then) 800-ish euro
20:55:00 <heat> aka not a high end thing
20:56:00 <heat> even had a mechanical drive lol
20:56:00 <zid> good thing about laptops is that you can buy ones with mech drives and crap ram
20:56:00 <heat> i'm not upgrading while CS2 doesn't force my hand
20:56:00 <zid> and then replace the drive and the ram
20:56:00 <zid> for cheeap
20:56:00 <ornx> huzzah, now everything works and the numbers make sense
20:56:00 <zid> 1 slot or 2?
20:56:00 <zid> 2x8GB is £22 on ebay
20:57:00 <zid> 1x16 is 24
20:57:00 <heat> 2 DIMM slots
20:57:00 <heat> sorry, SDIM? the laptop variant
20:57:00 <zid> SODIMM
20:57:00 <heat> oh yes, SODIMM
20:57:00 <zid> cheaper than the ddr3 I bought for my desktop before it died
20:57:00 <bslsk05> ​ SK HYNIX 16GB DDR4-2666V Laptop RAM HMA82GS6JJR8N-VK N0 AD 2Rx8 PC4-21300 | eBay
20:58:00 <zid> and I only got 8
20:58:00 <heat> cute, may consider but im not sure if its worth it anyway
21:00:00 <zid> I have.. no idea what this piece of junk has, it's too nice to care
21:01:00 <zid> okay I have two 900MHz ddr4s, and 4 slots, 16GBx2
21:55:00 <mcrod> hi
21:56:00 <gorgonical> I am losing my mind, but at least the "known working" solution is also wrong
22:32:00 <gorgonical> Who do I file a complaint with for hardware issues
22:32:00 <gorgonical> Who's tim apple but for arm
22:33:00 <zid> frank acorn
22:33:00 <klange> Rene Arm i guess, but you can also probably complain to Masayoshi Softbank
22:34:00 <gorgonical> At least I have ammunition that I'm not the idiot and the hardware is probably non-conforming
22:34:00 <gorgonical> I could still be the idiot but my case is stronger now
22:41:00 <gog> hi
22:41:00 <gog> it's me, janice x86
22:41:00 <zid> I am paul zilog
22:42:00 <gog> you're a pale copycat of me
22:42:00 <zid> I have autoincrementing loads
22:42:00 <gog> sksksksk
22:42:00 <zid> Is that you turning into a crab at last?
22:42:00 <gog> yes
22:43:00 <zid> It's like a magical girl power
22:43:00 <zid> but carcinization is the result
22:43:00 <moon-child> hey its me ur itanium
22:43:00 <heat> i'm richard spark
22:43:00 <zid> POWER OF CRAB!
22:43:00 <heat> sparc*
22:43:00 <zid> dick spark
22:43:00 <heat> dick sites
22:43:00 <klange> POWER OF WHALE
22:44:00 <heat> visit my domain at
22:44:00 <klange>
22:44:00 <bslsk05> ​'POWER. OF. WHALE!' by SkySophes (00:00:12)
22:44:00 <heat>
22:44:00 <bslsk05> ​playlist 'Dick Sites - Videos' by Dick Sites
22:44:00 <heat> dick sites is dicking sites
22:53:00 <heat> ChavGPT: srsly though what's so special about kutrace?
22:53:00 <heat> i don't get it
22:58:00 <kof123> mr. cowtools burroughs