IRC channel logs

<stikonas>hmm, I think struct siginfo_t is incorrect, that's why kaem fails to get non-zero exit codes

<stikonas>oriansj: https://github.com/oriansj/M2libc/pull/12 (this fixes kaem not reporting non-zero exit codes)

<oriansj>stikonas: M2libc merged

<stikonas>hex2 lseek is fixed now too, I'll fix hex1 tomorrow

<stikonas>and that should be it for riscv32

<oriansj>M2libc merged

<oriansj>and bootstrap-seeds merged

<stikonas>ok, I've pushed them to stage0-posix now

<stikonas>so that should be it for riscv32...

<oriansj>well make test-riscv32 seems to die at the first step

<stikonas>hmm, let me retry

<stikonas>hmm, indeed, must be something wrong in bootstrap seed

<stikonas>bootstrap-seeds/POSIX/riscv32/kaem-optional-seed kaem.riscv32 does work

<stikonas>oh, I think I know what's wrong

<stikonas>missed one jump in hex0 encoding

<stikonas>should be easy to fix, just need to move that initialization up in the file

<stikonas>yes, that fixes it...

<stikonas>sorry... Will update bootstrap-seeds then

<stikonas>ok, also fixed it in hex2 and M1 prototypes...

<stikonas>GAS prototype is missing for kaem-minimal... Not sure if we need it now

<stikonas>shouldn't be too hard to make it but right now I don't have riscv32 toolchain...

<stikonas>(except for M2-Planet, M1 and hex2)

<oriansj>bootstrap-seeds merged

<stikonas>ok, let me update stage0-posix

<stikonas>oriansj: ok, done

<stikonas>make test-riscv32 works now

<stikonas>would be good to have this tested on HW...

<stikonas>but at least in terms of registers we should be all initialized now

<oriansj>looks good

<stikonas>oh, I think there is another thing missing. error message in M2-Planet needs updating

<stikonas>wait, need to update tests...

<stikonas>hmm, probably intermittent error but I got the following

<stikonas> +> ./AArch64/artifact/M0 ./AArch64/artifact/M2-0.M1 ./AArch64/artifact/M2.hex2

<stikonas>qemu: uncaught target signal 11 (Segmentation fault) - core dumped

<stikonas>oh, it's only reproducible in qemu, works fine on real HW, probably caused by high memory usage

<stikonas>because M2-Planet has gotten bigger, we probably started hitting it more often in qemu now

<oriansj>M2-Planet merged

<stikonas>and stage0-posix updated too

<stikonas>also updated stage-posix in live-bootstrap

2021-12-04.log