Re: Kernel code and processor's Cache

On 9 Apr, 21:36, David Schwartz <dav...@xxxxxxxxxxxxx> wrote:
On Apr 9, 1:27 pm, feng <xu_feng...@xxxxxxxxx> wrote:

Cache dedicated to the kernel would slow the libraries.

I am still not convinced. if the Kernel code needs to  run in a system
with no  dedicated kernel cache, user application code and data can be
evicted from the cache to be brought back later. This will definitely
slowdown the execution of the user application. with dedicated kernel
cache, the content of the user cache is unchanged and hence user
application execution can be resumed without immediate cache Miss. How
can deciated cache to the kernel slow down user applications

Dedicated cache to the kernel would slow down the user applications
because they wouldn't be able to use that cache. On any realistic
system, 90% of more of the work is done by user applications. Taking
valuable cache away from them to make the kernel run faster would be
boneheaded.

DS

thanks David for your answer and especially your patience :). User
cache should still be there. i was just wondering why not ADD another
cache inside the processor along with a TLB dedicated solely to run
kernel routines. Kernel caching is done through all the memory
hierarchy apart from the processor...

Thank you
.

Relevant Pages

  • Re: Cached memory never gets released
    ... Stock linux 2.4.26 kernel. ... Due to flash bug 3M of memory gets lost due to font memory getting lost ... The output of "free" cache number steadily grows. ... longer to exhaust all of system memory with the cache. ...
    (Linux-Kernel)
  • Hung task - sync - 2.6.33-rc7 w/md6 multicore rebuild in process
    ... It seemed as though relcaim of fs cache became really slow once there were no more "free" pages. ... sync took a long time, and was reported by the kernel as a hung task - see below. ... # Device Drivers ...
    (Linux-Kernel)
  • Hung task - sync - 2.6.33-rc7 w/md6 multicore rebuild in process
    ... It seemed as though relcaim of fs cache became really slow once there were no more "free" pages. ... sync took a long time, and was reported by the kernel as a hung task - see below. ... # Device Drivers ...
    (Linux-Kernel)
  • Hung task - sync - 2.6.33-rc7 w/md6 multicore rebuild in process
    ... It seemed as though relcaim of fs cache became really slow once there were no more "free" pages. ... sync took a long time, and was reported by the kernel as a hung task - see below. ... # Device Drivers ...
    (Linux-Kernel)
  • Re: silent semantic changes with reiser4
    ... |>fixed ammount of space for disk cache is bad, ... kernel interface that then turns around and calls a userland daemon? ... one cache manager, the way there is exactly one VM manager for Linux. ...
    (Linux-Kernel)