Re: DWORD Alignment

Tech-Archive recommends: Repair Windows Errors & Optimize Windows Performance



"Thorsten Albers" <albersRE@xxxxxxxxxxxxxxxxxxx> wrote in message
news:01c57148$d095f640$a18ee684@xxxxxxxxxxx
>
> That is what the CPU returns. But AFAIK internally a 32 bit CPU gets
> always
> a 32 bit value but mapps it, if necessary, to the registers with the
> correct data byte width. But since I think I have read this somewhere and
> since my memory has like my money-bag some obvious leaks I wouldn't swear
> on this...

As far as I know, that would be all performance-related and would not have
an affect on function.

There are many possible perormance improvements that can vary depending on
the processor and other hardware. I think that most memory chips have a
cache of some type that can improve performance and the amount of memory
retrieved from memory and/or that is cached can vary. Processors have
internal caches. There might also be caches that store just 32 bits in 32
bit processors.

Most processors have machine instructions that are capable of loading a
register from the machine language statement. For example, in the following:

Dim ba(9) As Byte
ba(0) = ba(0) + 11

Since 11 is a constant, it could be translated into a machine instruction
that loads the Byte (8-bit) constant directly from the machine instruction
that is already in the instruction register.


.



Relevant Pages

  • Re: [PATCH] Document Linuxs memory barriers [try #4]
    ... +WHAT IS CONSIDERED MEMORY? ... and the division between CPU and memory needs to be marked out. ... Why do you need to include caches and queues in your model? ...
    (Linux-Kernel)
  • Re: Are prefix opcodes better than variable length?
    ... and the root cause of why Vector SuperComputers are no more. ... memory, so caches, which are primarily a latency improver, make sense. ...  So memory bandwidth hasn't improved as much as CPU cpeed. ...
    (comp.arch)
  • Re: anyone can help? weird Linux concurrency problem in math calculation
    ... I run two concurrent instances on a two-cpu Linux server, I'd expect that they'd take about the same time, but to my surprise, they took 31 seconds to finish. ... other existing jobs take negligible cpu times, ... no shortage of memory, ... means one might not benefit from the caches as much as one could. ...
    (comp.os.linux.development.apps)
  • Re: pcib allocation failure
    ... pcib1: attempting to grow prefetch window for ... attempting to grow memory window for ... cpu0: on acpi0 ... <ACPI PCI bus> on pcib0 ...
    (freebsd-current)
  • Next July 27: boot failure(hang) on x86_64 box.
    ... Freeing unused kernel memory: 1360k freed ... ACPI: PM-Timer IO Port: 0x488 ... CPU: L2 Cache: 1024K ... # AX.25 network device drivers ...
    (Linux-Kernel)