@lari @jaoler @jonne @osma @pesasa @rolle @vittorio_fossofino Sellaisen yleisen aaltoliikkeen oon vuosien aikana huomannut, että menneen high-level on nykypäivän low-level. Ihan ns. taviksetkin kirjoittelivat ysärillä AT-komentoja esim. pankkiasioinnin yhteydessä.

@lari @rolle @vittorio_fossofino @osma @pesasa @jaoler @jonne

#BBBS oli omasta mielestä mukavin, ja tiedonsiirtoprotokollista tietty Arisoft Oy:n #SMODEM. SMODEM oli askel #ZMODEM:sta siinä, että se mahdollisti kaksisuuntaisen tiedonsiirron purkkiin.

Purkkimaailman viimeisellä suoralla OS/2 Warp oli sysoppien keskuudessa kovassa suosiossa moniajo-ominaisuuksiensa vuoksi.

Joku on näemmä varastanut myöhempinä vuosina #Telemate:n tuotenimen, jota tuli käytettyä soitteluun. Pyhäinhäväistys mun mielestä :-) Eiköhän ne nimen käyttöoikeudet jne. ole jo rauenneet, eli tuskinpa sentään oikeussaleihin tällä keissillä pääsisi. Koitin kerran sähköpostilla tiedustella tästä kyseisestä firmasta kännissä ja läpällä vuoksi, mutta eivät vastailleet mitään.

Kaikki mitä purkkeihin soittelussa tuli opittua on ollut työssä koko työuran käytössä. Esimerkiksi AT-komennothan ovat Bluetooth-laitteiden kovaa ydintä, ja modeemithan niistä puhelimistakin edelleen löytyy. Tässä juuri siirtelen testiohjelman sisältävää tiedostoa FPGA:lla pyörivälle RISC-V-pohjaiselle CVA6-alustalle hyödyntäen ZMODEM-tiedonsiirtoprotokollaa :-)

I try to learn at least one RISC-V opcode a day to get more familiar. Today’s opcode was auipc which stores to a register a PC-relative address from an immediate offset. The offset is given as number of pages.

It is e.g. used to set the gp (which is mnemonic for x3), which is the base address for global variables.

Not trying to teach anyone. Writing things down like this just works for my memory :-)

#riscv #assembly #note

#china's non-nuclear ballistic missile plans are worst nuclear threat in ages...

remote working from here as i had to attend to a wedding over here... tonight back to #tampere :-)

i lost my #khachapuri #virginity at #tbilisi. delicious shit

@nirbheek One good thing to say about Apple (and actually also Amazon if you do the research but even more Apple): they don't cash you with your privacy nearly as much as other companies do.

That's why I use e.g. iPhone, I feel that I pay with money the whole device, rather than paying part of the device with euros and part with privacy. Also, all infosec researchers I've learned to know use iPhone for this exact reason. For better or worse, iOS-ecosystem is a castle.

If you compare prices of the early 00's GSM phones, the cheapest model iPhone SE is in the price range of phones such as Nokia 3210 when it was released. People often complain why iPhones are so expensive. IMHO, the more appropriate question why Android phones are dirt cheap...

eipä toi #vastaamo-todistusaineisto jätä hirveästi arvailun varaan. huvin ja harrastuksen vuoksi kiinnostaisi tietää, miten #KRP murti salauksen, mutta se jäänee pimentoon.

@uheplugins

@uheplugins https://tal-software.com/products/tal-dac

This #Behringer release made my realize what was missing from authentic #Roland JP-8000 sound when using @uheplugins #Diva: I need to add 12-bit bit reduction :-) With #ToguAudoLine #DAC added it hits the spot for me. https://www.youtube.com/watch?v=Trv8i3PyAAY&t=568s

@animatek As far as sound cards go I love my RME Babyface Pro FS :-) I boot it in class compliant mode when using Linux and proprietary mode when on macOS.

I need to really in-depth my RISC-V understanding and shift the focus away from the bug I've been struggling with (see various posts) :-) Reading all articles I can find related to CPU state changes, and also CVA6 user manual. So far have survived with trial and error but now I feel that I just don't have understanding of some bigger ideas... I should have taken my focus away from the issue at hand weeks ago but well you know...

@animatek #Pipewire has pretty much made the audio routing better than other operating systems. I use my Mac to do mostly music but when I’m on my work PC, I use Bitwig with Pipewire backend. I have Debian running on that.

With Pipewire I’m able to do stuff that I need Loopback on Mac, i.e. route audio somewhat freely between apps.

3rd party plugins that I actually use work too. For U-he stuff I have native Linux versions and FabFilter goodies I use through #yabridge.

Generally Windows plugins that are programmed with love in the first place (like FF stuff) works super solid with Yabridge, sometimes even more stable than real Windows. For super proprietary stuff like XLN Audio stuff I keep them only on my Mac.

Pipewire has sort of fixed things that I had to complain in the past with Linux audio.

I sort of like constructing tracks on my Linux PC given the more limited set of plugins. It keeps the focus on track construction, not on swapping plugins :-) I tend to finish the tracks on my Mac.

You should probably consider more realtime oriented kernel for the installation. I personally use https://liquorix.net/.

PS. I would not touch Ubuntustudio style distro’s because they tend to tune things sort of “old world” ways and sometimes even counter-productive. You are better off with e.g. standard Debian or Ubuntu installation. In addition to more real-time oriented kernel, you might want to add threadirqs to /etc/default/grub and install and systemctl enable rtirq with its default settings , This will give priority to your USB audio interface. I’m not sure if you even need at modern times to do anything to /etc/security/limits.conf given that by default Pipewire daemons do run already with real-time priorities and Bitwig has a Pipewire backend option.

To summarize, a stock distribution, real-time’ish kernel, rtirq and Pipewire would be me go-to list. At least I experience no glitches. And then the desktop is not tuned too heavily to a single purpose (audio), and configuration sort of stands better time as it is not too specific in its adjustments. Not that different from Windows or macOS in the end. In both you also need to few tweaks here and there to get a good result…

#CVA6 #PDF #manual: https://readthedocs.com/projects/openhw-group-cva6-user-manual/downloads/ #note

@nirbheek Actually also great for baremetal ARM testing since I can compile kernel images for RPi 3B+ without cross-compiler...

@nirbheek Made me pour almost 3k to my 32 GB RAM Mac Mini M2 Pro early this year so I guess it is still a working business model :------) Ya, I'm a bitch. It has been useful tho for native ARM64 (or at least "real" VM instead of emulation) kernel testing and music production.

@starshine password at any frequency pwgen -y 12 1

One constraint good to be aware of is that Keystone enclaves are always contiguous physical regions. Page tables are only tool for maintaining address space.

Have been passive at #LKML because I’m stuck with this page table sync bug in #Keystone: https://github.com/keystone-enclave/keystone/issues/378.

Snippet where it trips:

    sfence.vma
    csrw satp, a0 // switch to virtual addresssing
    sfence.vma

mtval is at the 2nd sfence.vma when misaligned store happens.

This never happens on QEMU and I presume that not on all hardware but I’ve tested this only with CVA6 (running on Genesys2 FPGA board). On QEMU everything just works.

At least, shouldn’t the sfence.vma be sort “useless” as it should continue where stvec is set, right? I guess here one thing to try out would be to simply remove the second sfence.vma?

I’m really just learning RISC-V sync opcodes and do not know e.g. when I should use also fence.i. Just trying different things without tbh knowing exactly what I’m doing yet…

In head.S there is also .align 2 after csrw and I have to admit that I don’t know exactly why… This bug is haunting me…