Hi! 21-Дек-2006 20:20 [EMAIL PROTECTED] (Japheth) wrote to freedos-devel@lists.sourceforge.net:
>>>> 1. On your site, program versions nowhere mentioned, only in "news" >>>> sometime. >> J> no, see http://www.japheth.de/Jemm386.html >> http://www.japheth.de/HX.html >> Find "2.1" -> not found J> Yes, but you claimed program versions are *nowhere* mentioned, so just 1 J> example will prove you wrong. Just add "_most_ programs". Here important not words (formal logic), but sense. >> Yes, skilled programmer may trace your makefiles (especially, there are >> two makefiles) and develop/modify to own one, which will fit into his >> environment, but this is _not_ "minimal" efforts. J> Aggreed, Then why not import proven build subsystem, which works at least for kernel and FreeEMM386? It should minimize your efforts to adopt different compilers (discussed subsystem is enough flexible for this) and it should be convenient for users (common for (some) programs subsystem; separation betwen settings and compile logic). J> it is *almost* "minimal" efforts. >> In yet only commented, but not removed. J> Yes, but I was in fear to be sued by the copyright holders (Tom Ehlert, ...) J> because of deleting their precious inline assembly code ... license.txt " 3. You may otherwise modify your copy of this Package in any way, provided that you insert a prominent notice in each changed file stating how and when you changed that file, and provided that you do at least ONE of the following: [...] c) rename any non-standard executables so the names do not conflict with standard executables, which must also be provided, and provide a separate manual page for each non-standard executable that clearly documents how it differs from the Standard Version. " >> Of course. But even skilled programmers (at least, me) slow down at >> such points and try to understand, whan happen. Especially, they/I should >> spend time to mentally translate, that "jnc" here mean "jae" (this >> translation is not most trivial task). J> I learnt assembly with the Z80, where there was no "jae" or "je", so as for J> me J> using jnc/jz are the most natural versions. (This is question of defining macros, if your assembler wasn't support relational mnemonics). Fortunately, x86 assembler allows better reflect programmer's intentions in given case - so, why not use these possibilities? Especially, not all was know z80. >> @@ok: >> xor ah,ah ; everything OK and finished >> ;clc >>>> [...] >> inc ax ; this is shorter, than "inc al" >> shl al,2 J> Indeed, saves a byte (although not true for "inc ax", since we are in 32bit J> mode), In 32-bit mode you may use "inc eax". Because AL <= 4, this is safe. J> but I deliberately choose to use "mov ah,00", because it clearly J> indicates that it is the "ok" return code which is set. For this, I (always!) use comments. Like this: xor ah,ah ; OPTIMIZE: XOR AH,AH instead MOV AH,0/CLC inc eax ; OPTIMIZE: INC EAX instead INC AL >> sub eax,edx >> jae @@noadj2 >> xor eax,eax >> @@noadj2: J> yes, this indeed is better. Go on! If you can reduce the size of V86 segment J> to 3400h it will save a full 4 kB page (currently size is still about 3540h) 320 bytes? Hard task, unless your code is ineffective :) or there is present some redundant functions. May you in short explain layout of your code? This should ease my findings. >> J> optimising the C part is irrelevant IMO >> Why not? J> It's used for initialization only. But it present in source and it present in executable itself. ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Freedos-devel mailing list Freedos-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/freedos-devel
