Jump to content

Talk:Address space

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Repurposing

[edit]

This shouldn't be just a redirect to IP address, should it? What about memory address spaces in computing? Keenanpepper 07:29, 3 Oct 2004 (UTC)

(This is now a real article. -- Beland 21:42, 28 November 2006 (UTC))[reply]

Harvard architecture

[edit]

Someone asked in an HTML comment:

Is 3 storage areas merely a type of Harvard architecture, or does "Harvard" imply exactly 2 storage areas ?
Having separate address spaces for instruction and data memory is what (in my opinion) defines the Harvard architecture. Assuming that there is only one instruction memory address space (I can't think of a reason to have more than one), that still leaves the door open for more than one data address space. This is common on digital signal processors, particularly thosed tuned to a specific purpose. For example, a hypothetical DSP highly optimized for FFTs might benefit from four data address spaces, real sample data, imaginary sample data, real twiddle factors and imaginary twiddle factors. So, my vote would be to call anything with one or more data address spaces distinct from the instruction address space Harvard.--GrahamDavies 13:27, 14 September 2007 (UTC)[reply]

Harvard: flat or not?

[edit]

This address space article currently claims:

Unfortunately, many early computers did not support a flat memory model -- in particular, Harvard architecture machines force program storage to be completely separate from data storage.

However, the (current version of) the flat memory model article contradicts that, saying that it is possible to have a Harvard architecture with a flat memory model, as long as it avoids segmentation and bank switching. I see that earlier versions of the "flat memory model" article[1] say that Harvard architecture is by definition not a flat memory model.

Can we resolve this contradiction? --68.0.124.33 (talk) 04:46, 9 September 2008 (UTC)[reply]

Diagram

[edit]

Added links to two diagram from commons illustrating LBA addressing and Virtual Memory Mapping. Removed reqdiagram tag, although a diagram about IP addressing would probably be good too. Egmason (talk) 02:41, 1 March 2012 (UTC)[reply]

Unfortunately I used that tag instead of starting a discussion here. My intention was to request/make a diagram about the abstract concept, not (prompt somebody) to search some stuff at the Commons. Certainly, I will remove these when true diagrams were made. Incnis Mrsi (talk) 04:43, 1 March 2012 (UTC)[reply]

The redirect Addressing has been listed at redirects for discussion to determine whether its use and function meets the redirect guidelines. Readers of this page are welcome to comment on this redirect at Wikipedia:Redirects for discussion/Log/2025 April 25 § Addressing until a consensus is reached. Duckmather (talk) 01:46, 25 April 2025 (UTC)[reply]

Unconsidered modifications

[edit]

I had made alterations of the precision and questions about the definition, which it looks be poor and imprecise.

Clearly, you can see these problems, however the user User:Guy Harris undid these alterations and it looks to be a Bounded rationality because not even the question about the possibly opinion-biased word “discrete”.What exactly does it mean in this context? An organization of the article in the Also section was undid too.

And it isn’t clear if this is limited to the RAM, ROM or both. If it’s just limited to that, this should be cited in the definition and the article could show the specifications about that.

Furthermore, I didn’t cite about RAM memory or something like that. “Address” is usually and POSSIBLY related to both, ROM or RAM memory. It isn’t limited only for RAM, only because the article speaks in the most part about the RAM memory, it doesn’t mean it limited for it, but if it is limited to the RAM memory, again, it needs to be defined on the start point.

For example, maybe a better definition could be:

In computing, an address space defines a range of discrete addresses, each of which may correspond to an user address space and kernel address space, network host, peripheral device, disk sector or other logical or physical address.[clarification needed]

instead of:

In computing, an address space defines a range of discrete addresses, each of which may correspond to a network host, peripheral device, disk sector, a memory cell or other logical or physical entity.

Also, if this concept is limited to the RAM, why isn’t the User Space and Kernel Space considered one of them? It’s contradictory.


Of course, I can be really wrong about the most of stuffs here. But, please, let's discuss about that. QuantumNinus (talk) 19:16, 9 January 2026 (UTC)[reply]

it isn’t clear if this is limited to the RAM, ROM or both It isn't limited to any of them. An "address space" can be a space of anything that's addressed - as it says, "each of which may correspond to a network host, peripheral device, disk sector, a memory cell or other logical or physical entity."
A network address "is an identifier for a node or host on a telecommunications network". The IPv4 network address of the network host on which I'm typing this, on the local area network I'm on, is 192.168.1.4, and the Wi-Fi MAC address of that host is f8:ff:c2:43:7d:c6.
The peripheral devices on that host have various addresses on various buses. Some of those may appear to the host's CPU(s) in the same fashion as memory addresses, in that instructions that can load data from or store data in memory can also read data from or write data to the registers for those peripheral devices.
That host has no disks, but it does have a solid-state drive, and each block on that drive has its own numerical address.
And its computer memory also has an address for each byte.
In computing, an address space defines a range of discrete addresses, In computing, an address space defines a range of discrete addresses, each of which may correspond to an user address space and kernel address space User space and kernel space says that "A modern computer operating system usually uses virtual memory to provide separate address spaces or regions of a single address space, called user space and kernel space.[1]" With some systems, those are separate address spaces; in others, they are parts of a single address space, so they shouldn't be referred to as "a user address space and kernel address space".
Also, if this concept is limited to the RAM It's not, even when not talking about those other types of addresses. A machine may have a physical address space, which is a range of computer memory addresses. The memory management unit and the operating system may support virtual addresses, and may provide to each process a separate virtual address space or may run all processes in the same virtual address space (a single address space operating system). The article needs to speak separately of physical addresses and logical/virtual addresses.
It should also stop talking about memory cells; it should talk about memory locations. Guy Harris (talk) 02:38, 10 January 2026 (UTC)[reply]
each of which may correspond to an user address space and kernel address space, Absolutely not; an address refers to a location within an address space, not to the address space itself. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 12:43, 11 January 2026 (UTC)[reply]
  1. ^ "Address space". Address space options for 32bit systems.