Return to Video

Sequence introduction cs348 unit7

  • 0:01 - 0:05
    We've solved the memory issue and we've solved the controlability issue.
  • 0:05 - 0:10
    However, there are more questions regarding traffic control. For example, what
  • 0:10 - 0:13
    if we had two input channels? How can we control the traffic streams on each
  • 0:13 - 0:18
    interface, separately? How can we synchronize between the traffic sent on one
  • 0:18 - 0:23
    versus the other channel? Is this scheme reusable? Can anybody understand this
  • 0:23 - 0:27
    scheme? You know, if you reuse code, can you just drop it on them, will they
  • 0:27 - 0:31
    find all the required things very easily? And also, can we constrain the
  • 0:31 - 0:36
    traffic depending on the state of the device? All these issues can be
  • 0:36 - 0:41
    individually solved by writing more code. However, it is not reusable. This is
  • 0:41 - 0:46
    why the developers of the first reuse methodology, ERM, decided to look at this
  • 0:46 - 0:50
    more closely, and develop a new reusable traffic generation scheme called
  • 0:50 - 0:57
    Sequences. Sequences work as follows. The traffic driver component is actually
  • 0:57 - 1:02
    separated into two subcomponents. The top one is called a sequencer and it
  • 1:02 - 1:07
    creates and randomizes traffic. The bottom component is the driver and he takes
  • 1:07 - 1:12
    the generated traffic and applies it to the interface. The driver is also
  • 1:12 - 1:17
    sometimes referred to as a BFM. which stands for Bus Functional Model. When the
  • 1:17 - 1:21
    driver is ready, it requests some traffic. The sequencer generates some
  • 1:21 - 1:25
    traffic, for example, a packet, and sends it to the driver which can then apply
  • 1:25 - 1:30
    the packet to the interface. The sequencer can also generate something else
  • 1:30 - 1:35
    which is called a Sequence. This is a struct which itself can contain lists of
  • 1:35 - 1:43
    packets and other sequences. So, this sequence contains a packet so it can send
  • 1:43 - 1:51
    it. But it also contains another sequence, which itself has packets. So, with
  • 1:51 - 1:55
    this setup, users can define what happens in a sequence, what kind of traffic
  • 1:55 - 2:00
    items are generated. Also, Sequences can be combined to create more interesting
  • 2:00 - 2:05
    scenarios. And finally, you can randomly pick a sequence to determine what kind
  • 2:05 - 2:09
    of traffic is sent to your device.
タイトル:
Sequence introduction cs348 unit7
Video Language:
English
Team:
Udacity
プロジェクト:
CS348 - Functional Hardware Verification
Duration:
02:09
Cogi-Admin added a translation

English subtitles

改訂