Re: Bug in usbser.sys?

Tech-Archive recommends: Repair Windows Errors & Optimize Windows Performance

From: Philip Lukidis (pagefault0x0_at_hotmail.com)
Date: 10/27/04

  • Next message: Shalini: "Re: HCT Tests Submission" 
    Date: Wed, 27 Oct 2004 16:33:26 -0400

    "Philip Lukidis" <pagefault0x0@hotmail.com> wrote in message
    news:uddlOe8uEHA.1300@TK2MSFTNGP14.phx.gbl...
    > "Randy Aull (MSFT)" <randyau@microsoft.com> wrote in message
    > news:417eef25$1@news.microsoft.com...
    > > Unfortunately I have never done any isochronous drivers so I don't know
    if
    > > this holds true there. I will check this out and get back to you.
    > >
    > > --
    > > This posting is provided "AS IS" with no warranties, and confers no
    > rights.
    > >
    > >
    > [snip]
    >
    > Thank you kindly. If what you posted does hold true for isochronous
    > transfers, please give me an indication of how to setup a multi-packet
    > transfer where one packet would not be of the same size (and this ideally
    > needs to work from 98SE-XP+...). Would the final packet in my buffer need
    > to simply be less than the maximum packet size?
    >
    > If what you have posted does not hold true for isochronous transfers,
    please
    > let me know for which type of transfers it does hold true for.
    >
    > thanks,
    >
    > Philip Lukidis
    >
    >

    Looking it over with the analyzer (it's in high demand here), and new WPP
    traces, I can confirm that this behaviour does not apply for isoch. I
    certainly receive a buffer composed of several packets less than the maximum
    packet size. Do the strictures which you have listed apply for interrupt or
    bulk transfers only?

    thanks,

    Philip Lukidis

  • Next message: Shalini: "Re: HCT Tests Submission"

    Relevant Pages

    • Re: Bug in usbser.sys?
      ... This posting is provided "AS IS" with no warranties, and confers no rights. ... "Philip Lukidis" wrote in message ... >> transfers, please give me an indication of how to setup a multi-packet ... Would the final packet in my buffer ...
      (microsoft.public.development.device.drivers)
    • Re: USB short packets and physical data packet length
      ... Is it more bandwidth efficient to simply define a second endpoint with a smaller wMaxPacketSize? ... USB transfers are organised as transactions in a microframe. ... The USB host arbiter will allocate timeslots for transactions based on the pipe type, maximum packet size of each pipe and the availability of data. ...
      (comp.arch.embedded)
    • Re: USB short packets and physical data packet length
      ... Interrupt endpoints, but not for Bulk or Control. ... or Control transfers are pending. ... During actual transfers, no reservation ... no packet is waiting, there is no slot reserved as these transfes are ...
      (comp.arch.embedded)
    • Re: USB short packets and physical data packet length
      ... Interrupt endpoints, the bandwidth is only reserved in each ... a "timeslot" doing nothing when Bulk or Control transfers are pending. ... I understand that if no packet is waiting, there is no slot reserved as these transfes are "on demand" and dynamically allocated, but I'd be very surprised if the USB arbiter worked on actual packet sizes and not max paxket lengths??? ... My argument was that small packet size would be more bandwidth efficient over a large packet size where the actual number of bytes was *always* smaller. ...
      (comp.arch.embedded)
    • Re: Bug in usbser.sys?
      ... I was just wondering if this holds true for all USB transfers types ... I/O request with a buffer of size 10*wMaxPacketSize, and say on the USB the ... how I can perform a transfer with different packet sizes in one I/O ...
      (microsoft.public.development.device.drivers)