ELF loading
in [ASM]
|
From: Betov on 27 Aug 2007 15:17 santosh <santosh.k83(a)gmail.com> �crivait news:fav4ac$fh6$1(a)aioe.org: > By SDL, do you mean this? > <http://en.wikipedia.org/wiki/Simple_DirectMedia_Layer> Yes. > But RosAsm does not need multimedia capabilities, does it? No. But SDL has some needed basic capacities, and is small. > How many different Win32 API calls does RosAsm contain? About 30 to > 40? 283. Called up to 301 times ('SendMessage'). > If you have spread out the Win32 calls all over the source, then it > will become very tedious to replace them. Yes, indeed. This is a scaring job to do. The API calls are not the only problem. > Just fork of the source and maintain a separate branch for > cross-platform use, (currently of course, just Windows and Linux), and > the usual Windows one. This way, you don't need to mess with the > existing stable version. > > To start with, you can disable the debugger and things like the PE > header generator. The actual instruction encoder shouldn't have to be > modified at all. So that leaves you with all the window, file and > memory management routines. If it was as simple as this... I have no idea where i will begin, for now. I first have to learn more about Linux. I am not even sure that something like RosAsm is simply possible to do under Linux. You have to consider the very huge number of volunteers for Linux, compared to the equivalent ones under Windows. If it does not exist, this is not a miracle. Why is Linux Assembly, to date, in the same state Win32 was, 12 years ago? Because .... there are problems. I know only a couple of those "problems". > Just use > GTK+. It's available even under Windows. QT is also available under > Windows. I have first to learn a bit of each... before thinking of making anything... > One alternative is to just write an ELF header generator, and to > disable the debugger. Then I think it can be immediately used under > WINE to produce ELF files. That's the easiest part. For day one. :)) Betov. < http://rosasm.org >
From: Betov on 27 Aug 2007 15:22 "rhyde(a)cs.ucr.edu" <rhyde(a)cs.ucr.edu> �crivait news:1188239317.952339.16160 @i38g2000prf.googlegroups.com: > having gone done the path of a proprietary source file format > myself (The LISA assembler, some 30 years ago), I can say without > hesitation, from experience, that this path is a mistake. Right: Keep sure that any step you could do would be in the wrong path, clown, as proven by the complete failure of your LISA adventures and HLA adventures. Well, as long as you make people laugh... :)) Betov. < http://rosasm.org >
From: santosh on 27 Aug 2007 15:54 Betov wrote: > santosh <santosh.k83(a)gmail.com> écrivait news:fav4ac$fh6$1(a)aioe.org: > > > By SDL, do you mean this? > > <http://en.wikipedia.org/wiki/Simple_DirectMedia_Layer> > > Yes. > > > But RosAsm does not need multimedia capabilities, does it? > > No. But SDL has some needed basic capacities, and is small. But are you prepared then to create your own widgets? Because SDL, as far as I can tell, provides no GUI widgets, just basic capability to output to screen. > > How many different Win32 API calls does RosAsm contain? About 30 to > > 40? > > 283. Called up to 301 times ('SendMessage'). Okay, this is a *big* problem then. > > If you have spread out the Win32 calls all over the source, then it > > will become very tedious to replace them. > > Yes, indeed. This is a scaring job to do. The API calls are not > the only problem. > > > Just fork of the source and maintain a separate branch for > > cross-platform use, (currently of course, just Windows and Linux), and > > the usual Windows one. This way, you don't need to mess with the > > existing stable version. > > > > To start with, you can disable the debugger and things like the PE > > header generator. The actual instruction encoder shouldn't have to be > > modified at all. So that leaves you with all the window, file and > > memory management routines. > > If it was as simple as this... I have no idea where i will begin, > for now. I first have to learn more about Linux. I am not even > sure that something like RosAsm is simply possible to do under > Linux. It should be because there are similar environments for HLLs. I can't think of anything possible under Windows that would not be possible under Linux. But, some functionality might be very difficult to duplicate though. > You have to consider the very huge number of volunteers > for Linux, compared to the equivalent ones under Windows. If > it does not exist, this is not a miracle. Why is Linux Assembly, > to date, in the same state Win32 was, 12 years ago? Because > ... there are problems. I know only a couple of those "problems". What are they? IMHO, the biggest dampner for assembly programming under Linux is the strong presence and culture of C. Nearly everything about Linux is implemented in C. Programmers who come to Linux from other UNIX systems tend to prefer a combination of C, Perl and shell scripting over anything else. I can't see any *technical* stumbling block for assembly programming in Linux. > > One alternative is to just write an ELF header generator, and to > > disable the debugger. Then I think it can be immediately used under > > WINE to produce ELF files. > > That's the easiest part. For day one. On actually trying RosAsm under WINE, I don't think so. WINE is not stable enough. BTW, is there a way to disable the TITLEs pop-up pane? It is not popping up properly for me, when I move the mouse cursor to the bottom edge. Is there a way to make it permanent, i.e. fixed and not popping up? Or alternatively make the whole source show up on the editor at once, not just the current TITLE.
From: Charles Crayne on 27 Aug 2007 16:33 On Mon, 27 Aug 2007 12:54:39 -0700 santosh <santosh.k83(a)gmail.com> wrote: > On actually trying RosAsm under WINE, I don't think so. WINE is not > stable > enough. I have run RosAsm under WINE, and, although I did not test it extensively, the only problem I encountered was with a permissions problem with the debug function. -- Chuck
From: Betov on 27 Aug 2007 16:54
santosh <santosh.k83(a)gmail.com> �crivait news:1188244479.224044.4200 @i13g2000prf.googlegroups.com: > But are you prepared then to create your own widgets? Because SDL, as > far as > I can tell, provides no GUI widgets, just basic capability to output > to > screen. I do not yet know. wxwidget = No. Fully custom interface = Too much work. Calling to Linux Interface = I don't yet know,... >> 283. Called up to 301 times ('SendMessage'). > > Okay, this is a *big* problem then. Anyway, yes. Having to port to another OS has never been planed, at all. I have bet everything on ReactOS, 10 years ago, at the risk of a complete failure. This failure is not yet 100% sure, but given what is going on, with MicroSoft buying the competition, and the little guys dreaming about the 20 millions dollars MS gave to LinSpire... we might be close to the failure. > It should be because there are similar environments for HLLs. I can't > think > of anything possible under Windows that would not be possible under > Linux. > But, some functionality might be very difficult to duplicate though. Sure. The Debugger will be the most difficult point. >> You have to consider the very huge number of volunteers >> for Linux, compared to the equivalent ones under Windows. If >> it does not exist, this is not a miracle. Why is Linux Assembly, >> to date, in the same state Win32 was, 12 years ago? Because >> ... there are problems. I know only a couple of those "problems". > > What are they? > > IMHO, the biggest dampner for assembly programming under Linux is the > strong presence and culture of C. Nearly everything about Linux is > implemented in C. Programmers who come to Linux from other UNIX > systems tend to prefer a combination of C, Perl and shell scripting > over anything else. The very first: The fact that *Linux* does not exist. You cannot write anything, in Assembly for an OS that does not exist. Now, if Ubuntu succeeeds to kill the competion, there is a hope. > I can't see any *technical* stumbling block for assembly programming > in Linux. Maybe, but fact is that Assembly does not exist, whereas it exists under Windows. Now, Linux Assembly *can* exist. NASM can do it. FASM can do it. But can something like RosAsm exist, under Ubuntu, this is far from sure. > On actually trying RosAsm under WINE, I don't think so. WINE is not > stable > enough. Yes, i also see strange effects. The first day i gave it a try, it seemed to me OK. Now, i see more and more problems. It works different each day. This give me the impression that this is me who is crazy. Very unpleasant, for somebody who probably is. :)) > BTW, is there a way to disable the TITLEs pop-up pane? No. > It is not > popping up > properly for me, when I move the mouse cursor to the bottom edge. Is > there > a way to make it permanent, i.e. fixed and not popping up? No. > Or > alternatively > make the whole source show up on the editor at once, not just the > current > TITLE. No. It would be too slow, anyway. The Editor is... special, because of its special features. Today, it flickers a lot, here... It now also fails to understand *where* it stands, which causes a re-install at each run. Well, strange indeed. Betov. < http://rosasm.org > |