@Doomed_Daniel@mastodon.gamedev.place avatar

Doomed_Daniel

@Doomed_Daniel@mastodon.gamedev.place

Linux User, Programmer, Gamedev, Metalhead ♠️ AD(H?)D ♠️ he/him ♠️ formerly on Twitter and mastodon.technology with the same handle

This profile is from a federated server and may be incomplete. Browse more on the original instance.

Doomed_Daniel, (edited ) to random
@Doomed_Daniel@mastodon.gamedev.place avatar

If anyone has a recent AMD (i)GPU and is running Linux, can you do me a favor and test something? Only takes one or two minutes:
$ wget https://github.com/dhewm/dhewm3/releases/download/1.5.3/dhewm3-1.5.3try2_Linux_amd64.tar.gz
$ tar xf dhewm3-1.5.3try2_Linux_amd64.tar.gz
$ cd dhewm3/
$ wget https://files.holarse-linuxgaming.de/native/Spiele/Doom%203/Demo/doom3-linux-1.1.1286-demo.x86.run
$ sh doom3-linux-1.1.1286-demo.x86.run --tar xf demo/
$ ./dhewm3

Once the mars globe in the main menu is at full size, on my Laptop (with Ryzen 7640U w/ RDNA3) I get these blue flickering specks; I wonder if everyone gets them.

Thanks in advance! :)

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

On my laptop this happens with Ubuntu 24.04 and also a very recent Arch Linux live image.

It does not happen on my old laptop with Intel GPU (with same Mesa and Kernel version), and not on my Desktop with nvidia GPU.
A friend couldn't reproduce it with his Radeon 5700 either.

But so far no one has tested it with RDNA3 (or even RDNA2), or AMD iGPUs in general (except for me on my Framework 13 AMD Laptop).
I wonder if this is a driver bug or maybe defective hardware or something.

Doomed_Daniel, (edited )
@Doomed_Daniel@mastodon.gamedev.place avatar

Doesn't happen on my secondary desktop with Radeon RX 580 (with the same Arch live system) either.

On the laptop the issue happens with X11, Wayland and XWayland.

Anyway, I'd be thankful about any reports, including "works on my system", ideally with information on your Linux kernel and Mesa version ($ glxinfo | grep "OpenGL renderer string") and which GPU/CPU your device has (for CPU: $ grep "model name" /proc/cpuinfo | head -n 1).

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@hugin thank you, good to know! :)

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@hugin
that's weird - did it print any relevant warnings to the terminal?

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@shinyshinken Thank you very much!

Steam Deck OLED would be very interesting, as it has a RDNA2 GPU, which is only one generation behind my RDNA3 GPU (and at least two generations newer than the Vega GPU in the 4800H)

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@GeorgePieVG
Thank you very much!

Doesn't the Ryzen 7 5800H have a GCN 5 (Vega) iGPU, which is the predecessor to RDNA1?

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@GeorgePieVG
no problem - but if you have a discrete AMD GPU (esp. one with RDNA), I'd be very interested if the bug happens there :)

lritter, to random
@lritter@mastodon.gamedev.place avatar

i'm increasingly growing fond of the API error handling style as it is done in GL 4; the idea being that you register a callback for things that went wrong, and then you can decide from there how to handle it: print a warning, log to file, abort (for debugger), longjmp, raise an error... it's more versatile than dealing with exceptions or return codes because you're still in the correct call stack for that error.

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@slembcke @lritter
those debug extensions have been around for a while now, both in desktop GL and OpenGL ES (at least with Mesa and nvidias driver?): https://www.khronos.org/opengl/wiki/Debug_Output

Note that to actually get your callback called from your offending gl* call (so you get a useful backtrace) you need to glEnable(GL_DEBUG_OUTPUT_SYNCHRONOUS)

Very useful! :)

distractal, to random
@distractal@mastodon.social avatar

https://www.piratewires.com/p/interview-with-jack-dorsey-mike-solana

This just made me a whole lot more interested in BlueSky.

Jack left because they started putting moderation tools in place and realized they needed rules, essentially.

What a dipshit.

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@distractal
It's still corporate social media with investors that (eventually) want to make money..
I see very little chance that it won't turn to shit eventually.

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@distractal
Also, didn't he already left BlueSky once several months ago?

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@distractal
I'm sure I read it back then, but because internet search sucks, the only sources I can find are https://www.reddit.com/r/BlueskySocial/comments/16l8cl7/why_dorsey_deleted_his_bsky_account/
https://www.reddit.com/r/BlueskySocial/comments/16jbbzr/jack_has_quit_bluesky/

Doomed_Daniel, to random
@Doomed_Daniel@mastodon.gamedev.place avatar

You know, on the one hand I totally understand that people don't want their StackOverflow answers to be used to train LLMs.

On the other hand, removing them or making them useless probably doesn't affect StackOverflow's LLM plans much (they just sell their data as of last week or whatever), but will make it even harder for actual people to find programming advice on the internet.

They may end up asking LLMs (and if they're lucky they get your old reply from SO :-p)...

Kadsenchaos, to random German
@Kadsenchaos@23.social avatar

Konservative sagen:

  • Recht auf Teilzeit abschaffen
  • Recht auf ganztägige Kitaplätze schwierig
  • Wochenarbeitszeit erhöhen
  • Wir brauchen mehr (deutsche) Kinder
  • Familien sind das Rückgrat der Gesellschaft

Was sie meinen:

  • Frauen an den Herd

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@c0c0bird @Kadsenchaos
Ich glaub dafür ist die CDU dann auch wieder zu geizig.

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@c0c0bird @Kadsenchaos
Endlich wieder Kinderarbeit einführen! Die FDP ist dabei.

wolfpld, to random
@wolfpld@mastodon.gamedev.place avatar

Gnome and window decorations, again.

https://factorio.com/blog/post/fff-408

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@dotstdy @pervognsen @wolfpld
reasonable question IMO

rovarma, to random
@rovarma@mastodon.gamedev.place avatar

The filesystem APIs on Linux remind me of Vulkan in a way. There's all these little low-level bits and bobs that can be used to build anything, but there's not a convenience function in sight, so everybody ends up implementing the exact same things in slightly different ways.

Sometimes a guy just wants to move a file to a different location, ya know?

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@rovarma @pervognsen
related: it's ridiculous that there's no copyfile() function

not just because it's annoying to implement it yourself, but also because it could do nice system-specific optimizations under the hood (on Linux >= 2.6.33 it could use sendfile(), on Linux >= 4.5 it could use copy_file_range(), some FSs support copy-on-write with reflink, and other Unices than Linux probably also have their own syscalls that could make it faster than copying yourself in userspace)

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@pervognsen @Donzanoid @rovarma @molecularmusing
Don't the "Windows Internals" books also have information on otherwise undocumented stuff?

Doomed_Daniel, to random
@Doomed_Daniel@mastodon.gamedev.place avatar

I haven't tried to reproduce that DHCP vulnerability, but as far as I can tell a good way to mitigate it could be to ignore those DHCP option type 121 settings from the server (which add static routes for IP subnets configured by the server, which could overwrite routes set by a VPN and thus circumvent the VPN) - unless those static routes are needed to reach your VPN endpoint in the first place, of course.

At https://mastodon.gamedev.place/ I wrote how this could be done on Linux.

Doomed_Daniel, (edited )
@Doomed_Daniel@mastodon.gamedev.place avatar

(... or other systems using the ISC DHCP Client)

By the way, it's remarkable how https://www.leviathansecurity.com/blog/tunnelvision doesn't even explicitly mention IPv4 vs IPv6 (though the examples are clearly v4, and RFC 3442 that introduced "option 121 classless static routes" is explicitly for DHCPv4).

I wonder if DHCPv6 (or even just plain IPv6 router advertisements or those ICMPv6 redirect messages?) can have the same issues?
Unfortunately I don't know enough about IPv6..

sos, to programming
@sos@mastodon.gamedev.place avatar

After 29 years, C++23 finally introduces "print" function.

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@sos
can the fugly iostreams finally die?

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@quelsh @sos
could be deprecated at least

pervognsen, to random
@pervognsen@mastodon.social avatar

The options in the classic sysmenu on Windows are still stuck in the Windows 3 era so it's nice to see features like this easily accessible for every window in KDE with a few clicks.

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@pervognsen
IceWM 20 years ago:

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@pervognsen
Yeah, definitely.
After getting used to this (and having multiple desktops, also visible on the screenshot behind 1, 2, 3, 4 on the taskbar), using Windows became (more) painful

BrideOfLinux, to random
@BrideOfLinux@mastodon.opencloud.lu avatar

"Android is the only operating system that fully immunizes VPN apps from the attack because it doesn't implement option 121. For all other OSes, there are no complete fixes. When apps run on Linux there’s a setting that minimizes the effects, but even then TunnelVision can be used to exploit a side channel": Novel attack against virtually all VPN apps neuters their entire purpose https://arstechnica.com/security/2024/05/novel-attack-against-virtually-all-vpn-apps-neuters-their-entire-purpose/

Doomed_Daniel,
@Doomed_Daniel@mastodon.gamedev.place avatar

@BrideOfLinux
Can't DHCP clients be configured to just ignore option 121?
For example, the ISC dhclient.conf has lets you overwrite options with "supersede" - not sure about the exact syntax, but something like supersede 121 ""; (or maybe the rulename, whichever it is, instead of 121) to set an empty rule might be possible?

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • anitta
  • thenastyranch
  • magazineikmin
  • tacticalgear
  • InstantRegret
  • ngwrru68w68
  • Durango
  • Youngstown
  • slotface
  • mdbf
  • rosin
  • PowerRangers
  • kavyap
  • DreamBathrooms
  • normalnudes
  • vwfavf
  • hgfsjryuu7
  • cisconetworking
  • osvaldo12
  • everett
  • ethstaker
  • GTA5RPClips
  • khanakhh
  • tester
  • modclub
  • cubers
  • Leos
  • provamag3
  • All magazines
    •