Posts by amonakov@mastodon.gamedev.place
(DIR) Post #B10sHeVmGm4rNmKeq8 by amonakov@mastodon.gamedev.place
2025-12-07T18:13:21Z
0 likes, 0 repeats
@wolf480pl oh, absolutely (I managed to think through some of that after posting)I'd expect bruteforce to find multiple preimage macs (a truncated hash is not a bijection), but yes, this is not an effective privacy measure
(DIR) Post #B18yCOkrZxHa3VwgwC by amonakov@mastodon.gamedev.place
2025-12-11T15:57:18Z
0 likes, 0 repeats
@wolf480pl did you try to follow what happens in case mac_randomize_bytes==3? it uses the kernel crypto api (AF_ALG socket) to hash a dozen of bytes, then reads three bytes of the hash into addr (leaving last three uninitialized), then wiphy_address_constrain overwrites the just-read three bytes with the initial three bytes of the hardware mac 🤦♂️ (end result: three-byte infoleak from stack)
(DIR) Post #B1PR3do0G2TTKOwwD2 by amonakov@mastodon.gamedev.place
2025-12-19T14:36:03Z
0 likes, 0 repeats
@wolf480pl @algernon at this point page faults on the giant maze array start to dominate, so you can get another 2x gain by outputting in a reasonable-size chunks
(DIR) Post #B1VerE1R7qwEf7imJs by amonakov@mastodon.gamedev.place
2025-12-22T14:38:55Z
0 likes, 0 repeats
@wolf480pl if you've seen my post about Брус-16 — pretty impressive what can be achieved in a compact designhave you played many games around CPU design? having an accidental "under complex microarchitectural conditions" moment in Silicon Zeroes was a highlight for me. I guess I should play Turing Complete eventually
(DIR) Post #B1Vg7Mgim2BM7uI9BY by amonakov@mastodon.gamedev.place
2025-12-22T14:53:05Z
0 likes, 0 repeats
@wolf480pl I'm actually not sure if Si0es is going to be very enjoyable with your kind of pre-existing experience, and in comparison to Turing CompleteI liked it when I played it when it was just released in 2017, but I feel in some ways its balance is questionable. To me it's more of a puzzle game than a sandbox game.
(DIR) Post #B1VgA2BFTddpOMqMAy by amonakov@mastodon.gamedev.place
2025-12-22T14:53:33Z
0 likes, 0 repeats
@wolf480pl it's here: https://mastodon.gamedev.place/@amonakov/115697045608891643
(DIR) Post #B1Vu8yi0IX9Th3YTbM by amonakov@mastodon.gamedev.place
2025-12-22T17:30:15Z
0 likes, 0 repeats
@wolf480pl I think this is an instance where brus-16 really shines as an educational platform, leaving such questions for the students to ponder and experiment with(e.g.: one possible next step is to recall that you have SUB, and with PUSHU + (ADD|SUB) you can shave off one more bit from PUSHU's immediate)
(DIR) Post #B1Vzg7IlrzCVfgNETw by amonakov@mastodon.gamedev.place
2025-12-22T18:08:57Z
0 likes, 0 repeats
I know I'm extremely many years late with this question, but... Is it just me, or is typical Docker workflow kinda wasteful? I just spawned a container with "podman run --rm -it alpine", installed two packages (+200 deps), found that the thing I needed to do is denied by seccomp, exited, relaunched with --security-opt "seccomp=unconfined"... and needed to grab those 202 pkgs from mirrors again.(we have a local cache/proxy to avoid fetches in CI jobs)
(DIR) Post #B1keXwJhdzoTcRT52e by amonakov@mastodon.gamedev.place
2025-12-29T20:09:27Z
0 likes, 0 repeats
I/O by chaining applications of operator() — like print("Hi, ")(username)('\n'), as in https://godbolt.org/z/jzMzva749
(DIR) Post #B1kenPVDPlqKTrEJd2 by amonakov@mastodon.gamedev.place
2025-12-29T20:18:57Z
0 likes, 0 repeats
@wolf480pl worse or better than chained applications of operator<< ?
(DIR) Post #B1kg4LTfjSXOsI9BgG by amonakov@mastodon.gamedev.place
2025-12-29T20:33:13Z
0 likes, 0 repeats
@wolf480pl aren't we allbut printing one thing should be print(thing)! and then the rest falls out from thatok, how about chained applications of operator, ? https://godbolt.org/z/KfnMd6qcz
(DIR) Post #B1kh4FSeCYBsAXBYvo by amonakov@mastodon.gamedev.place
2025-12-29T20:39:25Z
0 likes, 0 repeats
@wolf480pl or more lisp-flavoured https://godbolt.org/z/d1cnKbqTo
(DIR) Post #B1rsf2gZ7eM8raA7pw by amonakov@mastodon.gamedev.place
2026-01-02T07:47:00Z
0 likes, 0 repeats
@lanodan @wolf480pl I'd raise the stakes to "violates the x86-64 ELF psABI, which spells out the dynamic linker path very clearly" if it's elsewhere on Glibcbut musl ld.so does, in fact, avoid the name clash: https://gcc.gnu.org/cgit/gcc/tree/gcc/config/i386/linux64.hld.so paths for uclibc and bionic are here: https://gcc.gnu.org/cgit/gcc/tree/gcc/config/linux.h
(DIR) Post #B1s8hoQoAqYGn38Q3E by amonakov@mastodon.gamedev.place
2026-01-02T10:57:04Z
0 likes, 0 repeats
@wolf480pl oooh damn. I was going to jokingly ask if "going" was autocorrected from "GNU"I voted "perfectly reasonable" but that was assuming non-GNU Linux is in scope
(DIR) Post #B1x6FCjNWOt1dbqqRM by amonakov@mastodon.gamedev.place
2026-01-04T20:21:12Z
0 likes, 0 repeats
@asie can you say what the deal-breakers are? worse fragmentation in the simple allocator?
(DIR) Post #B25MeywlDysw05Q5vU by amonakov@mastodon.gamedev.place
2026-01-08T19:58:46Z
0 likes, 0 repeats
I just spent a few minutes looking up Kefir C compiler again after seeing it somewhere (probably someone linked to it here, thank you!) and forgetting what is was called or where it was (limiting the search to lobste.rs saved me). Also found that there's now a chibicc fork called slimcc: https://github.com/fuhsnn/slimcc
(DIR) Post #B2Hs6IezT2ItR28KA4 by amonakov@mastodon.gamedev.place
2026-01-14T20:53:28Z
0 likes, 0 repeats
@wolf480pl well, DVI overclocking is indeed a thing — I was running a FullHD monitor at 72 or maybe even 75 Hz instead of 60 for a while, using a EDID override (with reduced blanking intervals) to fit more frames in the 165 MHz limitI think it was heating up a little bit more, but otherwise worked as expected
(DIR) Post #B2RwfLwJKcM0sIM1x2 by amonakov@mastodon.gamedev.place
2026-01-19T17:24:13Z
0 likes, 1 repeats
Do any other CI runners have a feature similar to Sourcehut Builds that lets you easily ssh into a container with a failed job?
(DIR) Post #B2VYnnrnlIVRXSa7Cy by amonakov@mastodon.gamedev.place
2026-01-21T11:17:56Z
0 likes, 0 repeats
How am I seeing "length 11584" in tcpdump's packet summary on a bog-standard Ethernet with MTU 1500? For an inbound packet, that is.
(DIR) Post #B2n2veF9RTm9dd1a6q by amonakov@mastodon.gamedev.place
2026-01-29T21:41:26Z
0 likes, 0 repeats
LB: perhaps not widely known, but "stressapptest -W" is often a better diagnostic tool than memtest: detects not only memory but also CPU issues, triggers errors in seconds rather than minutes/hours, doesn't need a reboot.