Stories
Slash Boxes
Comments

SoylentNews is people

posted by cmn32480 on Wednesday May 23 2018, @06:47PM   Printer-friendly
from the your-computer-is-not-a-fast-PDP-11 dept.

Very interesting article at the IEEE ACM by David Chisnall.

In the wake of the recent Meltdown and Spectre vulnerabilities, it's worth spending some time looking at root causes. Both of these vulnerabilities involved processors speculatively executing instructions past some kind of access check and allowing the attacker to observe the results via a side channel. The features that led to these vulnerabilities, along with several others, were added to let C programmers continue to believe they were programming in a low-level language, when this hasn't been the case for decades.


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 2) by RamiK on Wednesday May 23 2018, @09:19PM (2 children)

    by RamiK (1813) on Wednesday May 23 2018, @09:19PM (#683283)

    Mill just got their fat pointers patents approved: https://patents.justia.com/assignee/mill-computing-inc [justia.com]

    Well, if it's any consolation, you can still justify parallel languages if they fail to solve micro-threading... But even if they miss the FPGA release this year, turfs should work well enough to do away with capabilities even with more conventional machines so the article isn't really going anywhere.

    Btw, when C++ introduced atomic operations and broke the memory model no one cared. So, it can argued that as long as the user land doesn't need substantial rewrites, new hardware requiring a new kernel will succeed commercially. Similarly, the way GPUs go 180 every few years and data-centers adapt further proves that as long as the required porting is localized user-land optimizations, hardware changes won't necessary lead to C being dropped.

    But hey, I still advocate Go and friends. But only because I think C sucks and we can and should do better.

    --
    compiling...
    Starting Score:    1  point
    Karma-Bonus Modifier   +1  

    Total Score:   2  
  • (Score: 2) by Wootery on Thursday May 24 2018, @01:22PM (1 child)

    by Wootery (2341) on Thursday May 24 2018, @01:22PM (#683533)

    The Mill guys are still going, huh.

    I see no mention of 'fat pointers' there. What are they?

    • (Score: 2) by RamiK on Thursday May 24 2018, @06:04PM

      by RamiK (1813) on Thursday May 24 2018, @06:04PM (#683652)

      Pointers with a bit extra around the waist :D So yeah it's just a general term for having more than an address. It could be range resulting in a bounded pointer... It could be ownership resulting in a capability pointer... It could have meta bits signifying explicit data type... Typically it's a mix of the above.

      The Mill patents in question covers their variation (turfs), roughly how it's implemented at the CPU and MMU levels and how they've circumvented the problems the article raises as reason to abandon the C memory model.

      --
      compiling...