Eero Tamminen dixit: >I wasn't thinking that Hatari would be used for running the system, >just help in debugging issues in getting it running.
Right. >Issues where Hatari can help are unlikely to be dpkg related. My idea was more to the point of testing the MMU, which would, I think, involve doing some sort of workload. >Hatari's main advantage over Aranym are more accurate (030) emulation >and its debugger. > >I think Hatari would be best at investigating: >- bootup problems >- crashes that cannot be investigated at C-level >- performance bottlenecks, mainly kernel because kernel it's > at fixed address Right, I can see its strengths there. >All together the user-space processes take about 4MB of RAM. >Then there's the amount of memory used by the kernel on top >of that. Oh, okay. >Are there some processes at bootup which dirty MBs of memory >at bootup and which quit before login (as I saw no such thing), >or which memory map very large files? I don’t think so. >If not, I think bootup should fit well into 14MB (ST-RAM). :-) Right. On the other hand, I heard rumours that Linux needs TT-RAM to function at all, and that ST-RAM wasn’t available to applications. That’s another reason I wanted the mailing list in the loop, to confirm, deny or possibly fix that ;-) >> The installation process is especially memory sensitive, as >> it involves a ramdisk of lots of megs… > >Could you point to documentation describing it more in detail? There’s none really, but you have a kernel and an initrd, and the entire process runs from RAM. Wouter even did build d-i images, but they’re not yet usable for various reasons. >Linux memory usage analysis is one of my specialties. Oh. Great, good to know! Thanks, //mirabilos -- Solange man keine schmutzigen Tricks macht, und ich meine *wirklich* schmutzige Tricks, wie bei einer doppelt verketteten Liste beide Pointer XORen und in nur einem Word speichern, funktioniert Boehm ganz hervorragend. -- Andreas Bogk über boehm-gc in d.a.s.r -- To UNSUBSCRIBE, email to debian-68k-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/pine.bsm.4.64l.1301221253010.20...@herc.mirbsd.org
