|
The value of assembly language programming lies in the ability to access hardware bits directly, and to execute CPU instructions that are not available to applications written in HLLs. A device driver is one example of an ASM application.
Few programmers have a real, in-depth understanding of how the hardware works, primarily because they have little need for such knowledge in their work. Assembly programmers must have an intimate knowledge of the hardware in order to write effective, safe code. While I believe that most programmers would benefit from more in-depth knowledge, it isn't really necessary for writing applications. Learn ASM for your own reward - you may find that it's fun to manipulate bits and flags in the CPU directly - but don't plan on writing applications with it. That's just not practical.
"...putting all your eggs in one basket along with your bowling ball and gym clothes only gets you scrambled eggs and an extra laundry day... " - Jeffry J. Brickley
|
|
|
|
|
Dear Roger, Thank you very much for your comment. Could you give me more idea from here?[^]
A thousand mile of journey, begin with the first step.
APO-CEDC
Save Children Norway-Cambodia Office
|
|
|
|
|
I have a problem while launching windows update feature (w2k). The message is:
Windows Update has encountered an error. This may be due to a discrepancy in your computer's time setting.
To check your date and time setting:
On the taskbar, double-click the time.
Verify that the date and time is correct.
I can't get to the update page then.
Well, I don't have time setting problem; my time is correct.
Can anybody suggest a solution, please?
Regards
|
|
|
|
|
I got the same message yesterday on Windows Server 2003 RTM. There may be a bug.
Stability. What an interesting concept. -- Chris Maunder
|
|
|
|
|
There was a bug, indeed. After some time it worked ok. I suppose, Microsoft team was adding a untested feature to this site.
|
|
|
|
|
liquid_ wrote:
I can't get to the update page then.
Well, I don't have time setting problem; my time is correct.
Can anybody suggest a solution, please?
I used to heard my teacher told me about an error with the windows time zone. According to his suggestion, he told me to change the windows time zone to any time zone that is difference from the default time zone after we install windows 2000. He told me that, that is the recommend from Microsoft that need the user to do that everytime when they finish the installation process of windows 2000. Whether your time zone is correct with the default time zone that originally install by windows, but Microsoft still ask you to change to any difference time zone, then apply it back to the time zone that is the default time zone after it install windows 2000. But he did not tell me more detail about the behavior of an error or show me the document that he told me.
A thousand mile of journey, begin with the first step.
APO-CEDC
Save Children Norway-Cambodia Office
|
|
|
|
|
I am trying to access my computer at home from the computer at my office.Both are Turned on.But telnet is not responding.It gives the error message can't be connected at port 23.Can anyone solve this problem and please tell if there is some specific format.
Be FaithFull To Your Work.
|
|
|
|
|
That will occur when youe home pc has no telnet server/port opened.
|
|
|
|
|
Your home computer has to have the Telnet service running, and if you are using a firewall program, it has to have the Telnet service and port enabled. Usually this option must be accompanied by a list of authorized users and IP addresses, so be prepared when you attempt to enable it to provide that information.
"...putting all your eggs in one basket along with your bowling ball and gym clothes only gets you scrambled eggs and an extra laundry day... " - Jeffry J. Brickley
|
|
|
|
|
Usman Tasleem Akshaf wrote:
It gives the error message can't be connected at port 23.
In my opinion, you did not enable the Telnet service in the Services management consol. Try to enable it and try your connectivity again. On the other hand, if you need to learn on how to use telnet[^].
A thousand mile of journey, begin with the first step.
APO-CEDC
Save Children Norway-Cambodia Office
|
|
|
|
|
Hi,
I am creating a batch file which will display current folders in the given directory with there creation date.The date i am getting is mm/dd/yy format only using the following:
set folder_date=%%~td
I want the format like mm/dd/yyyy.How do i get this format ?
Any help...
Thank you.
|
|
|
|
|
Go to control panel, open Regional and Language Option. On the Regional Options tab click on the Customize... button. The Customize Region Option dialog box appear and click on the Date tab. Choose the specific format of Short date format combobox and apply the setting. The new format will apply in your command prompt.
|
|
|
|
|
I'm developing an application running on Windows CE .Net 5.0 and sometimes I get the message 'intoofueil'. This will occur each second for a long period of time. When setting the unit into SystemIdle_L2 it will stop.
Does anyone know what this means and is it harmfull (besides filling my debuglog file with a lot of these messages).
|
|
|
|
|
Hi Expert,
Does anybody know how Win2K works exactly of recognizing the date format ? I got difficulties that my remote DCOM is recognizing the date format in the server as MM/dd/YYYY while in the regional setting of that PC I set to dd/MM/YYYY ?
Furthermore I found an article in MSDN:
http://support.microsoft.com/default.aspx/kb/271587
suggesting to set the
[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\OLEAUT]
VarConversionLocaleSetting = 2
to force using system default setting, but it doesn't work (after I did all they suggest, by setting that registry & the system default to dd/MM/YYYY, my remote DCOM still getting MM/dd/YYYY).
Is there something else I need to do ? (I don't use IIS, so I didn't set anything releated to IIS).
Thanks before.
|
|
|
|
|
hi,
everything about the compilers always stops at the creation of an assembly lang file and being fed to the assebnler.
how does the assembler translate?what exactly is a binary data( know it is voltage level etc) that can be given directly to the processor from keyboard?
the lack of info is killing me.someone please help.i want to know.
bye.
|
|
|
|
|
as i don't understand what you talk about exactly, i certainly won't answer the whole things you want to know.
When you program (whatever language), the compiler has to transform plein text (source code) to binary datas.
binary datas is a queue of bytes which are understood directly by the microprocessor.
processor manufacturers design built-in functions such as additions, decimal division, long jumps, etc... this way, software must fit to that built-in functions for them to be executed (else, they won't be even understood - or misunderstood).
these built-in functions are recognized by the processor as one ore two consecutive bytes, followed directly by its parameter(s).
the goal of the compiler is to translate the source code into that machine code made of bytes ones behind others...
of course, machine code are processor dependant and before the compiler starts working, it must get the processor type (intel,IBM, amd, etc...) to generate the right language.
if you want to know the processors assembly commands and byte translation, you will have to look in processors manuals.
when you talk about voltage levels, i'm not certain to understand you. if you talk about how to transmit the generated bytes to the processor physically, it is certainly in the proc manual. but that job is done by the operating system.
cheers,
TOXCCT >>> GEII power
[toxcct][VisualCalc]
|
|
|
|
|
Actually few modern compilers bother with an intermediate assembler stage, unless you actually ask for a listing. They go straight to object code.
A program as an executable file on disk isn't really runnable as-is. It's more like flat-pack furniture - all the parts are there (in theory!) but some assembly is required. The tool that assembles the program into a running memory image - a process - is the loader, a component of the OS. This creates the process, sets up the address space, loads the executable and all its dependencies (relocating libraries as required), performs the linking from import to export tables so that calls to DLLs work properly, creates a thread stack, then a thread, sets the initial thread's entry point to the address listed in the executable's header, and finally switches the processor to that thread, causing it to start executing.
What loads the loader? Various bits of the OS load each other; the OS starts loading from the disk's boot sector, which the BIOS locates and loads. The BIOS itself is run at system startup because it is electrically wired to a range of physical memory addresses starting at the point that the processor is designed to read its first instruction from.
When dealing with hardware there are two main methods of receiving input. The first is simply to read the device periodically to see if there's any data - known as polling. Polling is pretty inefficient because you have to read frequently enough to remain responsive. For example the human optical system cannot detect things that take less than about 0.04 seconds to occur - if it takes less than this it is observed as being instantaneous. You therefore need to poll at least this fast in order to give the illusion of instantaneous processing. But if the user isn't pressing any keys, it's wasteful - you could have spent that time doing something else.
In order to poll the processor must be able to read from the device. The I/O devices are therefore wired up so that when particular addresses appear on the address bus, the I/O device responds to the read request. (Some processors, like the x86 family, have a separate I/O line to distinguish I/O operations from memory operations, and corresponding IN and OUT instructions to cause this line to have a different value when using IN or OUT from that used when performing memory operations. Other processors have memory-mapped I/O, i.e. I/O devices just appear as if they're a kind of memory).
So most modern input devices use interrupts. The processor has some inputs on it known as interrupt requests (IRQs). If it detects that an interrupt is occurring, it jumps to a known location associated with that interrupt, and executes the code at that location. When processing is complete the interrupt handler uses a special return-from-interrupt instruction to continue whatever task was interrupted.
Traditionally the processor was the only device in the system that generated addresses on the address bus. All transfers from an I/O device to or from main memory had to be done explicitly in software - normally a loop that read from an I/O location (either using IN or a memory-read operation) and wrote to memory, or vice-versa. This is known as programmed I/O. Again this takes processor time that could be used for other things.
Later devices support Direct Memory Access, using bus arbitration lines on the processor to cause the processor to stop outputting addresses onto the bus, then drive the address and data buses itself. In the older architectures this caused the processor to 'stall', but with caches the processor can continue working even when it doesn't have access to main memory. The processor bus is normally a lot faster than the other buses, so the device normally only steals a relatively few cycles anyway. In this mechanism the OS simply instructs the device where to transfer data from or to, the device gets on with it, then signals via an interrupt when the operation is complete.
Stability. What an interesting concept. -- Chris Maunder
|
|
|
|
|
I like the answer that Mike gave you, but it needs a bit more detail to fulfill what you're asking. No matter what level of language you use to attack a problem, at some point the solution has to be translated into a pattern of ones and zeroes that a specific CPU can interpret and execute. That's the assembler's job. Each CPU on the market has a unique instruction set, which is defined as a set of binary numbers which cause the CPU to execute specific instructions. In reality, the binary number is an address which points to a starting point in the CPU Control Store that contains the binary switching information necessary to execute the specified instruction. In practice, what we see is an opcode, plus 0 or more operands contained in memory.
For example, for the old Intel 8080 CPU, a Jump instruction was defined as the hex value 0xc3, followed by two bytes of address data. When the program counter found the byte 0xc3, it then looked for two bytes of address data in the next two contiguous bytes of RAM, then transferred execution to the instruction at that address. The Assembler program takes care of filling that memory address with the proper binary data by using #define statements to assign specific values to variables, as in
#define Target 0x100
JMP Target
which would cause program execution to jump to the instruction located at 0x100.
Assembly programming requires an intimate knowledge of the target hardware and its internal structure. High level languages relieve us of that requirement, and make life much easier by abstracting the details of implementation so that we no longer have to worry about which registers are available, and whether memory addressing is high-Endian or low-Endian (Intel quirks). Using high level languages makes life a whole lot easier for all of us except those who have to worry about the details.
Voltage requirements are beyond the need-to-know threshold for programmers. If you can get the proper ones and zeroes into memory, we hardware types will take care of the rest. I strongly believe that a programmer who understands how the target hardware works will be a better programmer, but I think that I'm in a small minority group here - you decide. But I'm proud of you for wanting to understand. No knowledge is ever wasted...
"...putting all your eggs in one basket along with your bowling ball and gym clothes only gets you scrambled eggs and an extra laundry day... " - Jeffry J. Brickley
|
|
|
|
|
Hi all,
does WinXP provide a way to automate "Defrag"? It could be done by creating an script and schedule it to be executed at a certain time but, in a computer running 24x7, the best way should be to run defrag in background mode when the resources load is low or when there's no HDD requests from the applications. I know that there are some commercial products that perform that but I'd like something provided by WinXP itself ... or free???
Thanks,
Marc Soleda
|
|
|
|
|
Never heard about automatic defrags..
Don't try it, just do it! 
|
|
|
|
|
|
Thanks, I'll keep in mind to apply it another time
... she said you are the perfect stranger she said baby let's keep it like this... Dire Straits
|
|
|
|
|
You are wellcome 
|
|
|
|
|
We are creating a simulation environment where the user will be creating input data sets for hundreds of individual 3D atmospheric simulations. Each individual simulation takes between 1-3 hours to run to completion on a 3GHz machine. The simulation code is old Fortran executables invoked from Python scripts.
The computing environment is a "loose" cluster of ordinary PC's running Windows/XP Pro with a few off-the-shelf 1 Gigabit 10BaseT switches to connect it all together. One machine shares a large data disk to all the other "cluster" members to provide a common rooted file system.
The current plan is to use Condor as the Batch engine.
Does anyone on this board have any opinions or thoughts about this configuration??
Cheers... Martin
|
|
|
|
|
I have an e-mail 2 mega and i want to make it 250 mega . can you help me ?
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
