All about Security, News and Incidents for Linux OS

Discussion in 'Linux Compatibility and Software' started by Papusan, Oct 11, 2021 at 9:34 PM.

  1. Papusan

    Papusan Jokebook's Sucks! Dont waste your $$$ on Filthy

    Likes Received:
    Trophy Points:
    If the thread title could be improved, then please add in suggestions. Thanks:) @jclausius +++

    :vbbiggrin::vbbiggrin::vbbiggrin::vbbiggrin::vbbiggrin: :vbbiggrin:

    Linux Developers Push Urgent Patches To Fix 'Yet Another Hardware Trainwreck' | Monday, October 11, 2021

    Accurate timers are critical to the function of the low-level parts of the underlying code that drives the user-facing software we actually use in our daily lives. Fortunately, x86-64 PCs include numerous timers. Actually selecting which of those timers to use in a given scenario, however, can be a headache due to bugs, design flaws or implementation issues.

    In essence, when modern Intel chips hit the PC10 idle state, the HPET stops working, even if the operating system is currently using it. This, obviously, results in all kinds of problems, so the solution is to fall back to older and less precise timers, like PMTIMER.
  2. jclausius

    jclausius Notebook Virtuoso

    Likes Received:
    Trophy Points:
    Yes. Timers are extremely important. And the hardware will let you know if you can support a Real Time Operating System (RTOS). I know with Win32's API they do expose different timers based on what you're wanting the timer to do and how accurate it needs to be (think video editing vs. a by-the second alarm clock - each will have different accuracy).

    I'll have to dig into the article to see where programmers are becoming confused, but it wouldn't be surprising that people become confused by which hardware clock is used for what. I wonder if AMD chips have the same issues with timers??

    Papusan and Vasudev like this.

Share This Page