Re: [00/17] Large Blocksize Support V3

On Thu, Apr 26, 2007 at 01:37:42PM +0100, Andy Whitcroft wrote:
Nick Piggin wrote:
Christoph Lameter wrote:
On Thu, 26 Apr 2007, Nick Piggin wrote:
What are the exact requirement you are trying to address?

Block size > page cache size.

But what do you mean with it? A block is no longer a contiguous
section of memory. So you have redefined the term.

I don't understand what you mean at all. A block has always been a
contiguous area of disk.

Lets take Nick's definition of block being a disk based unit for the
moment. That does not change the key contention here, that even with
hardware specifically designed to handle 4k pages that hardware handles
larger contigious areas more efficiently. David Chinner gives us
figures showing major overall throughput improvements from (I assume)
shorter scatter gather lists and better tag utilisation.

I haven't actually provided any figures - it's knowledge passed down
from those that know more about it that I do.

If you want figures about the impact of large I/Os, then we should
not be looking at the HBAs but at the impact on RAID controller
throughput (this I do have numbers on ;). It is not uncommon to
see 2MB I/Os give twice the throughput of 512K I/Os to a single
RAID controller - larger than 512k can only be acheived on systems
with a page size larger than 4k.....

I am loath to
say we can just blame the hardware vendors for poor design.

Never did - I'm pointing out that linux can't use all the
capabilities they have.

Cheers,

Dave.
--
Dave Chinner
Principal Engineer
SGI Australian Software Group
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Relevant Pages

  • Re: hyper threading.
    ... If you know the design ... of the hardware, you can predict that two processors will provide x% ... generally improve throughput and should never diminish it (disregarding ... the difference between 100% busy and less than 100% busy, ...
    (freebsd-questions)
  • Re: Matlab code and MIPS
    ... I am new to hardware implementation from Matlab code. ... But if someone can help me how do we estimate or caculate MIPS from the existing Matlab code. ... Throughput measurements can be made ...
    (comp.soft-sys.matlab)
  • Re: FC2 + newest openoffice = slower
    ... months old (Athlon XP-M 2800+, 512meg ram, new HD). ... he manages to get 20MB/sec throughput with his changes. ... Most interestingly, despite the MUCH newer hardware I'm running, my throughput ... Even with the same settings he uses. ...
    (Fedora)
  • Re: how can I decrease the time cost when synthesis and implement
    ... productive in weeding out the bugs than trying to debug on actual hardware. ... Running the same test in hardware may cost an hour in synthesis but it then allows viewing the whole video sequence in real-time instead of having to wait a few days after the simulation. ... Throughput is one good argument for hardware testing. ... With hardware validation as a backup, simulations can be more focused on small, highly feature-specific tests while hardware runs help identify missing test cases and problematic areas based on real-world results. ...
    (comp.arch.fpga)
  • Re: Throughput
    ... actual hardware that you need? ... Also do you mean 100 Mbps (megabit per ... second) throughput via the external interface? ...
    (microsoft.public.isa)