Announcement

Collapse
No announcement yet.

The "Memory Folios" Work Continues - Improving Linux Performance, 7% Faster Kernel Builds

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Garp
    replied
    Originally posted by nsneck View Post
    Interesting and impressive work. Of course looking forward to many more patchsets, with this maybe landing next year.

    Why does this require separate patches to XFS, and why has this only been done for XFS? From the patches this seems to touch just the kernels memory management side, do filesystems really need separate logic for new MM paths?
    Both the author of the memory folio work, and the core maintainer for XFS, work for Oracle. XFS is a nice and mature kernel component, probably makes for a logical proof of concept.

    Leave a comment:


  • arQon
    replied
    A few years back, a NAS I worked on was using an ARM chip that could work with either 4K or 64K pages. Between the slowness of the CPU (~600MHz), and the overhead of the massive page tables needed to hold GbE-worth of data in-flight to/from an HDD, the change to the 16x larger pages netted something like a 15% increase in performance, finally allowing the woefully-underresourced thing (it had basically half the RAM it "should have" had) to be able to actually sustain line rate rather than constantly drooping (sic) down to 800-ish Mb/s thanks to all the extra work the 4K build had to do.

    So while this really only "matters" for the ends of the spectrum - i.e. weak ARM devices and powerful servers - rather than any typical desktop case, it genuinely can matter for at least those extremes, and this patchset is very much a Good Thing in its own right, even without counting the code cleanup etc.
    Last edited by arQon; 17 June 2021, 09:15 PM.

    Leave a comment:


  • Vistaus
    replied
    Originally posted by CochainComplex View Post

    Hhhhehehehehh TP hehehe
    Oh, grow up.

    Leave a comment:


  • CochainComplex
    replied
    btw...

    Brilliant Guy: writing this great patchset on memory folios
    Linux Geek Community: ....."bungholio....hehehehehhhee TP !"
    Brilliant Guy: ....

    Leave a comment:


  • jabl
    replied
    Originally posted by FireBurn View Post

    Don't worry, the author was clearly struggling too
    Here's a short and clear overview what it's about: https://lwn.net/Articles/849538/

    Leave a comment:


  • CochainComplex
    replied
    Originally posted by skeevy420 View Post
    Memory Folios.....THP....struct folio

    I am the Great Cornholio. I need TP for my bunghole.

    Yeah, this is one of those articles that's just a little above me. Somewhere around the middle it turns into "Well, at least those are words in English".
    Hhhhehehehehh TP hehehe

    Leave a comment:


  • jabl
    replied
    Originally posted by nsneck View Post
    Interesting and impressive work. Of course looking forward to many more patchsets, with this maybe landing next year.

    Why does this require separate patches to XFS, and why has this only been done for XFS? From the patches this seems to touch just the kernels memory management side, do filesystems really need separate logic for new MM paths?
    Filesystems are tightly integrated into the workings of the page cache, so yes, there is a close coupling between filesystems and the MM code. I haven't looked at the code, but I would guess that the XFS changes amount to working with 'folios' instead of individual pages. As for why XFS in particular, I guess that's just the fs the developer is most familiar with. It doesn't make sense to convert every filesystem yet before the core part is accepted upstream.

    Leave a comment:


  • FireBurn
    replied
    Originally posted by skeevy420 View Post
    Memory Folios.....THP....struct folio

    I am the Great Cornholio. I need TP for my bunghole.

    Yeah, this is one of those articles that's just a little above me. Somewhere around the middle it turns into "Well, at least those are words in English".
    Don't worry, the author was clearly struggling too

    Leave a comment:


  • boxie
    replied
    Originally posted by nsneck View Post
    Interesting and impressive work. Of course looking forward to many more patchsets, with this maybe landing next year.

    Why does this require separate patches to XFS, and why has this only been done for XFS? From the patches this seems to touch just the kernels memory management side, do filesystems really need separate logic for new MM paths?
    I would suggest that is more that any kernel work needs a user (i.e. not adding new things without having something use it). If they can make XFS (which I am guessing is their default FS) faster, then that's a double win

    Leave a comment:


  • skeevy420
    replied
    Memory Folios.....THP....struct folio

    I am the Great Cornholio. I need TP for my bunghole.

    Yeah, this is one of those articles that's just a little above me. Somewhere around the middle it turns into "Well, at least those are words in English".

    Leave a comment:

Working...
X