|
It's basically whether the segment selector is used first, or the page number is used first. See Segmentation and Paging Combined[^].
Note, I found this through the obvious method: Google for '"segmented paging" "paged segmentation"'.
In practice, segment selectors that are not part of a 'flat' virtual address, such as x86 processors have, are a total pain in the neck to set up and use - which is why Windows ignores them for the most part, only using them as the processor requires. Most processors now operate on a segmented paging basis, where the 'page table' is actually a hierarchy of page directories and page tables.
The original 386 scheme uses the most significant 10 bits of the virtual address to access the top level (the Page Directory) which contains a physical address pointing to the appropriate Page Table. The next 10 bits are used to index the appropriate Page Table, which then contains the physical address of the actual page. On Pentium and later processors, the page directory entry contains a bit (Page Size) which indicates whether the remaining 22 bits in the address are a page table index and offset within the page, or whether they are an offset into a large page (4MByte). Windows 2000 and XP use 4MByte pages to map parts of system address space, to reduce the number of page tables and TLB space, which gives more memory available to programs, at the cost of higher internal fragmentation.
If you enable Physical Addressing Extension, the format of the page tables changes. There are now three levels: Page Directory Pointer Table, Page Directory, and Page Table. The top level (PDPT) contains four Page Directory Pointer entries, then the Page Directory and Page Table each contain 512 entries. Each entry is 64 bits in size, some of which point to an address and some of which are flags. Again, the Page Table level can be omitted, with the Page Directory Entry pointing directly to a 2MByte page. AMD's AMD64 architecture adds a fourth level for Long Mode (64-bit virtual addressing).
The advantage of these multi-level systems is that not all page directory entries need be valid - the OS need only construct and fill in the page tables that are actually being used. Indeed, it can swap out page tables if required.
|
|
|
|
|
Can anyone tell me a way to access my solaris server at school using my windows machine at home. What i need is the complete GUI . I already use PUTTY and Telnet.
chao,
Ram
|
|
|
|
|
|
How can I remove the admin shares in Windows XP Professional SP1a using a code source or some value in registry setting?
Thanks
"Don't let me be missunderstood"
|
|
|
|
|
|
Hi !
Me and a classmate are working on a degree project for our university. We are working on a program that identifies a user in the Active Directory on a Windows Server 2003 by its fingerprints.
We are newbies when it comes to programming with ADSI but so far we have manged to write programs that can create and alter user and search the AD for users with a custom "hasFingerPrints"-attribute flag that we have created. We are using Visual C++ 6
The final program will be used when an application want the current user to verify its identity.
When the user enrolls the program creates one or several binary template-files (depending on how many fingerprints the user registers) with encrypted information about the fingerprint.
The fingerprint template files have a minimum size of 1kb but can be up to 4-8 kB if you enroll with a higher security level ... each user might also have 3-5 files ... one for each finger he enrolls.
One problem is that the current identification algorithm needs the file to exist in a directory of some form ...
What is the best way to store these files ?
Should they be stored inside the AD-database ? for maximum security if so, how do you do that ?
It would be great if there was a way to "attach" each fingerprint file to each user ... somehow
Thanks in advance !
|
|
|
|
|
Yes I have Googled for this but all I get is 5 billion hits from sites trying to sell me the RAM I already have.
I have a stick of Kingston KTD-GX150/256 3.3 v CE SD-RAM. It says it is Equivalent to Dell RAM. Is this only useful for Dell machines or is it OK to use in any SD-RAM machine as all SD-RAM is 3.3 V and the same as any other?
Michael Martin
Australia
"I suspect I will be impressed though, I am easy."
- Paul Watson 21/09/2003
|
|
|
|
|
It's okay to use it in any motherboard that has the same specs. Most of the proprietary memory standards that mfrs impose are far more strict than those used by clone makers, and you usually pay a premium price for them. I recently had a customer who needed to boost the RAM in her HP Laserjet printers. At the HP website it costs $300 per printer to upgrade; on the open market, the same RAM module costs $26. I doubt very much that to HP logo is worth that much, since the RAM is generic PC100 RAM.
"Your village called -
They're missing their idiot."
|
|
|
|
|
Roger Wright wrote:
It's okay to use it in any motherboard that has the same specs.
By that do you mean any motherboard that takes SD-RAM or any motherboard that mentions 3.3 V explicitly?
<EDIT>
A couple of more things I have found out tonight. This 256MB stick and another 128MB PC133 SD-RAM stick would not work in a Dell OptiPlex GX100 running a Celeron 433MHz at 66MHz. I assume because the RAM is 2 steps above the requirement, though this hasn't stopped me using PC133 in a GX1 Pentium II a couple of years ago.
I am hesitant to put the Kingston RAM back in my Athlon machine as last time I put it in with the 512MB PC133 SD-RAM already in there I had major problems. Windows started fine then crashed. Then on subsequent reboots I had incorrect memory readings at POST and got wavey staticy lines all over the screen during POST. Eventually the whole installation of XP sh*t itself even after I took out the RAM. Last 5 - 6 weeks since the rebuild it has been rock solid. So son't want to tempt fate though want the extra memory.
Excerpt from my motherboard (GA7-VKMLS) manual on memory -
- 2 168-pin SDRAM DIMM sockets
- Supports PC-100/PC-133 SDRAM (Auto)
- Supports up to 1GB DRAM (Max)
- Supports only 3.3V SDRAM DIMM
- Supports 64bit DRAM integrity mode
Can't understand why I am having the RAM problems I am.
</EDIT>
Michael Martin
Australia
"I suspect I will be impressed though, I am easy."
- Paul Watson 21/09/2003
|
|
|
|
|
In a single PC you want to avoid mixing types - you shouldn't run PC100 with PC133. The motherboard attempts to detect which one you're using and then sets the buss clock rate accordingly, so a mix will confuse it. Putting fast RAM in a slow motherboard can also cause problems, as dynamic RAM requires a periodic refresh to maintain charge in its cells and the slow board may not refresh often enough. The voltage is critical; 5 volt RAM won't work, or will work unreliably in a 3.3V system, and 3.3V RAM in a 5V system will smoke.
Some manufacturers do odd things with memory; I've had Micron systems refuse to operate with anything but Micron RAM. I know that to call it PC100 or any other such industry standard name they have to meet a certain spec, but I think some of them interpret the spec differently. If that's the case, you're stuck. The info in your manual sounds like they intend it to work with any 3.3V PC100/133 RAM, but if Kingston designed this stick to match some weird, non-standard feature of Dell's you may have trouble with it. I'd make a backup of the system and give it a go in any case.
"Your village called -
They're missing their idiot."
|
|
|
|
|
I'm actually leaning towards using the Kingston RAM (Dell equivalent) in the Dell Optiplex GX100 but even that won't play ball with the single 256MB stick in.
The Dell manual says that it handles 256MB sticks so it should work. Also at the office I have a couple of these machines running the 64MB PC100 stick that they came with along side a generic 128MB PC133 stick without missing a beat for the 6 or so weeks they have been in. Something is just trying to f*** me over and it's giving me the sh*ts.
I want to turn this Dell box into a RedHat Linux 9 box that will act as a Web and Mail Server for me as well as a PDC for my Windows boxes. Just for use inside my home network mind you, and I can't even get the RAM to work in it.
I wish these f***ers would all stick to the standards created for the hardware.
Michael Martin
Australia
"I suspect I will be impressed though, I am easy."
- Paul Watson 21/09/2003
|
|
|
|
|
"File permission canonicalization exploit attempt detected"
I received it from my firewall a few minutes ago (in fact, it's still being blocked) and I'm curious what it's describing. The intrusion came from a university in Sao Paulo, so I assume it's a student playing on Sunday, not the professor this IP block is assigned to.
"Your village called -
They're missing their idiot."
|
|
|
|
|
|
Thanks, Michael!
"Your village called -
They're missing their idiot."
|
|
|
|
|
The "File Permission Canonicalization" is one of the most simple and basic forms of attack to a web server. Basically, you add ".." to the URLs and try to get to the root of the drive of the web server.
Most crappy web servers have this problem, and IIS had it too on the past.
If you see these requests on your web server log:
/scripts/..%5c../winnt/system32/cmd.exe
/msadc/..%5c../..%5c../..%5c/..
/scripts/..%5c../winnt/system32/cmd.exe
GET /MSADC/root.exe HTTP/1.0
Probably it's just the Nimda worm.
Trying to make bits uncopyable is like trying to make water not wet.
-- Bruce Schneier
By the way, dog_spawn isn't a nickname - it is my name with an underscore instead of a space. -- dog_spawn
|
|
|
|
|
I've seen a couple of those - thanks for the clarification!
"Your village called -
They're missing their idiot."
|
|
|
|
|
I'm going nuts - no email for two days! My ISP is clueless, so I've been doing some spelunking on his network myself trying to find a possible cause. Ping reports that it can't resolve the name of the mail server, leading me at first to believe that his DNS is down. Nslookup also fails to locate the ISP's DNS server, though it does locate the server upstream from the ISP on the espire.net system. But even this one reports that it can't find the local mail server. On the other hand, I have no trouble surfing the 'net, so DNS must be responding. This leads me to conclude that the local DNS is dead, the mail server is down or not responding, and that it has been dead long enough that the upstream DNS MX record has expired. I'm going across the river to drop off a payment at their office in a couple hours, so I'll pass on what I've found if they're in the office, but if there are any other tests I can run to further isolate the problem I'd like to know about them. Any suggestions?
"Your village called -
They're missing their idiot."
|
|
|
|
|
Oh well, the bonehead finally got it to work, after a long, 18-hour shift. If he'd bothered to call I'm sure that this community could have solved the problem in about 40 seconds... Thanks anyway, guys and gals...
"Your village called -
They're missing their idiot."
|
|
|
|
|
I was born intelligent
Education ruined me!.
|
|
|
|
|
Confused about what? When the ISP I use breaks, the admins (if they're having a smart day) call me and I usually find something that they've missed or done wrong - it's almost always helpful to get an outside opinion before you drive yourself batty trying to unscrew the inscrutable on your own. If I can't figure it out, sometimes I submit the problem to the fine geniuses here at CP for help, and they generally come up with a solution. Other times, when I really don't care whether my connection is working that day, I just tell them I have no idea and let them squirm for a few hours or days.
My guess is that one day, whoever owns the place will figure out that he should have hired me to run it instead of the cable monkey he's currently using.
"Your village called -
They're missing their idiot."
|
|
|
|
|
How can i use this exe for remote computer? I always have error:access denied if i used parameter "/s remotecomputername". Should the remote computer sharing WINDOWS folder? Please help me.. Thanks
|
|
|
|
|
1) You must be able to ping the other computer.
2) It would require a username and a password.
For more information follow this link: http://www.microsoft.com/technet/treeview/default.asp?url=/technet/prodtechnol/winxppro/proddocs/tasklist.asp
|
|
|
|
|
Ok, i'm on xp. I have the admin account on a computer. and students use the "Students" account, with "limited privilidges".
my question is... how can I make a folder, and give the "Students" privilidges to write to that folder, but.... not delete anything from it, or delet the folder it self. BUt, still allow the admin to delete stuff.
any tips?
bye the way... this is all on a local hard drive. not over the network or anything... on a singel computer..
thanks.
/\ |_ E X E GG
|
|
|
|
|
I don't think you can, but I am probably wrong!!
BTW if you have write access, to "delete" a file all you need to do is just overwrite it with garbage.
"The greatest mistake you can make in life is to be continually fearing you will make one." - Elbert Hubbard
|
|
|
|
|
Have you tried the following:
1) Right click on the directory itself and go to "Sharing and security"
2) Select "Share this folder"
3) Press the "Permissions" button.
4) Give everyone, Read permissions and yourself, Full Control.
I am thinking that this would prevent anyone else from deleting any content of tha directory or that directory itself, but will let you do whatever you want with it.
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
