CPUID - характеристики центрального процессора

...
  • New test.

    g1.jpg
    First window - codename bug fixed, ext to e fixed, but second std is now in contact with number.
    Second window - PDCM - no/yes is not showed at all. Some design issues with places of no, yes in 2nd and 4th column.
    Beginning of txt tifle - codename is saved now. Also I don't like the change made by Powerdoge with the highest std and ext calls - the decimal numbers are shown instead of hex ones.

    g2.jpg
    Checking text file. The strange thing - I don't see SSE 4.1 in txt.file. All new features saved in some random order, so it is hard to check. I'll do it later.
    Attachments
    g1.jpg
    g1.jpg (180.84 KiB)
    Viewed 11608 times
    g2.jpg
    g2.jpg (181.08 KiB)
    Viewed 11608 times
  • Updated version. Text is file is in random order.
    Attachments
    FINALoNE.png
    FINALoNE.png (117.92 KiB)
    Viewed 11584 times
    trunkFinal.zip (182.22 KiB)
    Downloaded 305 times
  • I'll check it soon. Also, we need a new exception message Unknown model (maybe Unknown Codename), like we have messages
    'SORRY, CPUID IS NOT AVAILABLE'
    'SORRY, THIS VENDOR IS NOT SUPPORTED YET'
    'Next generation CPU'
    'EXTENDED FEATURES ARE NOT AVAILABLE'
  • Sure. I will add that.
  • So, the results of check - the following issues since midterm were fixed 1) redraw bug, 2) abstract examples in visualisation block, 3) duplication in txt file.
    And merge of beta2007 was done.
    But since 2007 some things changed in features, so you have to check specifications about changes. We already found, that 3DNow! abandoned or reduced, but probably show positive on check (and was in CPU's for years), related instructions like 3DNowPrefetch should be checked also, SSE5 flag was replaced by 3 others flags (you need to check this...) .
  • Any news?
  • I've got files for checking, but I need help from someone with AMD processor....

    Perspectives - dump of registers to the same text file, where others things are saved.
  • pre-release version 2.43 for testing with binary and sources inside
    Attachments
    test.zip (94.66 KiB)
    Downloaded 303 times
  • KolibriOS was run in VirtualBOX
    Attachments
    cpu-z.png
    cpu-z.png (17.69 KiB)
    Viewed 11436 times
    cpuid.png
    cpuid.png (17.82 KiB)
    Viewed 11436 times
    "Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius -- and a lot of courage -- to move in the opposite direction." Albert Einstein
  • So, we have:
    1) a lot of design issues in 2nd and 4th strings, including some label on top of Codename string.
    2) incorrect extended family... Probably, I fixed 3 years ago calculation for Intel, but not for AMD.
    3) incorrect Codename. May be we should do branching by family <> ext family and do special detection of new AMD CPU's as we do for new Intel CPU's
    4) we should disable multiplier detection for new AMD in the same way as we do it fo Intel
    5) L3 cache was detected by CPUID, but not in CPU-Z...
    6) AMD-V technology - we still dont' add support for its detection.

    Side note - we should change design of one of blue buttons, so it would match the other one.
  • How? People. Manage to do. Such. Horrible. Photos.

    Besides, about font...
  • dunkaist,
    your report shows mainly the same things + new one
    7) SSE5 support is a surprise, since SSE5 doesn't exist https://en.wikipedia.org/wiki/SSE5 , so something in the same bit detected.
    We still have feature detection on the level of September 2007.
  • Who is online

    Users browsing this forum: No registered users and 2 guests