<davidlt[m]>
Conan Kudo: kernel.org finally got 6.0.10. 6.0.11 already got a couple of minor riscv changes queued.
<davidlt[m]>
Oh, that reminds me that I should make kernel-ark PR.
zsun has joined #fedora-riscv
davidlt has quit [Ping timeout: 260 seconds]
davidlt has joined #fedora-riscv
zsun has quit [Ping timeout: 264 seconds]
zsun has joined #fedora-riscv
<davidlt[m]>
Rust is really fighting with me. 1.64 failed again, but only at packaging stage.
<davidlt[m]>
1.65 is still waiting for it's turn. 1.66 might be back to normal situation (but problem not fixed).
<davidlt[m]>
Somehow I get impression that Intel/SiFive HorseCreek will feel so fast that it will be hard to comprehend it. Yet, it will be slow compared to anything else.
<davidlt[m]>
I cannot expect much from years old IP design considering how fast things are progressing.
<davidlt[m]>
P550 was succeeded by P650, which was succeeded by P670 (minor, but some compliant to RISC-V Profiles [which is not ratified yet]).
<davidlt[m]>
I constantly wonder what we will do if/when RVA23 becomes a major profile, and none of announced core IP is RVA23 compliant yet. Well, RVA23 doesn't even exist in any draft form, just RVI slides.
<davidlt[m]>
I do expect majority of standards based RV platforms to be RVA23 compliant, but there will be a very short period of RVA22 (SiFive and Andes already announced it).
<davidlt[m]>
One could also expect that existing RVA22 might be RVA23 compliant too (if/when that gets defined).
<davidlt[m]>
I am looking at SiFive P670 brief on their website and I do see vector and vector crypto, which will be required (if nothing changes) in RVA23. Support for non-4K pages will also be required (most likely).
<davidlt[m]>
But that's not listed in the brief.
zsun has quit [Quit: Leaving.]
davidlt has quit [Ping timeout: 260 seconds]
cyberpear has quit [Quit: Connection closed for inactivity]