<html>
  <head>
    <meta content="text/html; charset=utf-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <div class="moz-cite-prefix">Hi Oded,<br>
      <br>
      Is the r600g driver ready for the Gallium3D infrastructure of big
      endian hosts or did you only solve the wrong colors problem?<br>
      <br>
      Cheers,<br>
      <br>
      Christian<br>
      <br>
      On 04 August 2015 at 6:57 PM, Oded Gabbay wrote:<br>
    </div>
    <blockquote
cite="mid:CAFCwf11ZmfDp=4KnCUKXn=3YsG5L+t5mHB4_mTkNj0z8rBBhdw@mail.gmail.com"
      type="cite">
      <div dir="ltr">
        <div class="gmail_default" style="font-size:small"><br>
        </div>
        <div class="gmail_extra"><br>
          <div class="gmail_quote">On Mon, Aug 3, 2015 at 6:40 PM, Emil
            Velikov <span dir="ltr"><<a moz-do-not-send="true"
                href="mailto:emil.l.velikov@gmail.com" target="_blank">emil.l.velikov@gmail.com</a>></span>
            wrote:<br>
            <blockquote class="gmail_quote" style="margin:0 0 0
              .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi Oded,<br>
              <span class=""><br>
                On 2 August 2015 at 11:37, Oded Gabbay <<a
                  moz-do-not-send="true"
                  href="mailto:oded.gabbay@gmail.com">oded.gabbay@gmail.com</a>>
                wrote:<br>
                > This patch fixes a bug that is manifested in the
                read path of mesa when<br>
                > running on big-endian machines. The effects can be
                seen when running<br>
                > piglit sanity test and/or taking a screen capture.<br>
                ><br>
                > The bug is caused when _mesa_format_convert
                receives src_format as<br>
                > mesa_format, which it thens changes to
                mesa_array_format. During this<br>
                > change, it checks for endianness and swaps the
                bytes accordingly.<br>
                > However, because the bytes are _already_ swapped in
                the memory itself<br>
                > (being written there by llvmpipe), and src_format
                value matches the<br>
                > _actual_ contents of the memory, the result of the
                read is wrong.<br>
                ><br>
              </span>I'm assuming that you're looked at swrast +
              softpipe as well - do they<br>
              use the same approach or is llvmpipe the odd one out ?<br>
              <br>
            </blockquote>
            <div>
              <div class="gmail_default"
                style="font-size:small;display:inline">​Hi Emil,</div>
            </div>
            <div>
              <div class="gmail_default"
                style="font-size:small;display:inline">​</div>
               </div>
            <div>
              <div class="gmail_default" style="font-size:small">​I
                checked it with swrast, softpipe AND llvmpipe.</div>
              <div class="gmail_default" style="font-size:small">Without
                my patch, all methods fail piglit sanity on ppc64</div>
              <div class="gmail_default" style="font-size:small">With my
                patch, all ​methods pass piglit sanity</div>
              <div class="gmail_default" style="font-size:small"><br>
              </div>
              <div class="gmail_default" style="font-size:small">   Oded</div>
              <br>
            </div>
            <div> </div>
            <blockquote class="gmail_quote" style="margin:0 0 0
              .8ex;border-left:1px #ccc solid;padding-left:1ex">
              Curious if your work has any effect on the big endian +
              r600 issue<br>
              [1]. I believe Christian Zigotzky (Cc'ed) was very
              passionate about<br>
              getting his r600 working - perhaps he can give your
              patches a test ?<br>
              ... just to make sure that things don't go even worse :)<br>
              <br>
              <br>
              Cheers,<br>
              Emil<br>
              <br>
              [1] <a moz-do-not-send="true"
                href="https://bugs.freedesktop.org/show_bug.cgi?id=72877"
                rel="noreferrer" target="_blank">https://bugs.freedesktop.org/show_bug.cgi?id=72877</a><br>
            </blockquote>
          </div>
          <br>
        </div>
        <img moz-do-not-send="true" class="mailtrack-img"
src="https://mailtrack.io/trace/mail/e9a7eba62edf3d06bbd1e78730aa6efb45954b43.png"
          height="0" width="0"></div>
    </blockquote>
    <br>
  </body>
</html>