<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html;
      charset=windows-1252">
  </head>
  <body>
    <p>Hum, 4294967295 = 0xffffffff = max uint32 value. This really a
      suspicious file size that looks like truncation at the 4 GB limit.
      Are you really really sure that you tested the non-GDAL
      implementation with that exact file, and not another version that
      might have been truncated later ?<br>
    </p>
    <div class="moz-cite-prefix">Le 07/04/2021 à 14:59, ni hao a écrit :<br>
    </div>
    <blockquote type="cite"
cite="mid:DM5PR02MB38628A04E03E4A902261FF12F1759@DM5PR02MB3862.namprd02.prod.outlook.com">
      <div>
        Hi Even,</div>
      <div>
        <br>
      </div>
      <div>
        I am using GDAL 2.4.4. I may try the latest version as you
        suggested.</div>
      <div>
        <span><br>
        </span></div>
      <div>
        <span>The NITF file size is 4,294,967,295 bytes </span></div>
      <div>
        <br>
      </div>
      <div>
        <span>The vendor used a non-GDAL implementation, and was able to
          display all image.</span><br>
      </div>
      <div>
        <span><br>
        </span></div>
      <div>
        <span><span>Metadata:</span>
          <div>
              NITF_ABPP=16</div>
          <div>
              NITF_BLOCKA_BLOCK_COUNT=01</div>
          <div>
              NITF_BLOCKA_BLOCK_INSTANCE_01=01</div>
          <div>
              NITF_BLOCKA_FRFC_LOC_01=</div>
          <div>
              NITF_BLOCKA_FRLC_LOC_01=</div>
          <div>
              NITF_BLOCKA_LAYOVER_ANGLE_01=000</div>
          <div>
              NITF_BLOCKA_LRFC_LOC_01=</div>
          <div>
              NITF_BLOCKA_LRLC_LOC_01=</div>
          <div>
              NITF_BLOCKA_L_LINES_01=19694</div>
          <div>
              NITF_BLOCKA_N_GRAY_01=00000</div>
          <div>
              NITF_BLOCKA_SHADOW_ANGLE_01=180</div>
          <div>
              NITF_CCS_COLUMN=0</div>
          <div>
              NITF_CCS_ROW=0</div>
          <div>
              NITF_CLEVEL=07</div>
          <div>
              NITF_ENCRYP=0</div>
          <div>
              NITF_FBKGC=  0,  0,  0</div>
          <div>
              NITF_FDT=</div>
          <div>
              NITF_FHDR=NITF02.10</div>
          <div>
              NITF_FSCATP=</div>
          <div>
              NITF_FSCAUT=</div>
          <div>
              NITF_FSCLAS=U</div>
          <div>
              NITF_FSCLSY=CA</div>
          <div>
              NITF_FSCLTX=</div>
          <div>
              NITF_FSCODE=</div>
          <div>
              NITF_FSCOP=00000</div>
          <div>
              NITF_FSCPYS=00000</div>
          <div>
              NITF_FSCRSN=</div>
          <div>
              NITF_FSCTLH=</div>
          <div>
              NITF_FSCTLN=</div>
          <div>
              NITF_FSDCDT=</div>
          <div>
              NITF_FSDCTP=</div>
          <div>
              NITF_FSDCXM=</div>
          <div>
              NITF_FSDG=</div>
          <div>
              NITF_FSDGDT=</div>
          <div>
              NITF_FSREL=</div>
          <div>
              NITF_FSSRDT=</div>
          <div>
              NITF_FTITLE=</div>
          <div>
              NITF_IALVL=0</div>
          <div>
              NITF_IC=NC</div>
          <div>
              NITF_ICAT=SAR</div>
          <div>
              NITF_ICORDS=G</div>
          <div>
              NITF_IDATIM=</div>
          <div>
              NITF_IDLVL=</div>
          <div>
              NITF_IGEOLO=</div>
          <div>
              NITF_IID1=P1</div>
          <div>
              NITF_IID2=</div>
          <div>
              NITF_ILOC_COLUMN=0</div>
          <div>
              NITF_ILOC_ROW=0</div>
          <div>
              NITF_IMAG=1.0</div>
          <div>
              NITF_IMODE=P</div>
          <div>
              NITF_IREP=MULTI</div>
          <div>
              NITF_ISCATP=</div>
          <div>
              NITF_ISCAUT=</div>
          <div>
              NITF_ISCLAS=U</div>
          <div>
              NITF_ISCLSY=CA</div>
          <div>
              NITF_ISCLTX=</div>
          <div>
              NITF_ISCODE=</div>
          <div>
              NITF_ISCRSN=</div>
          <div>
              NITF_ISCTLH=</div>
          <div>
              NITF_ISCTLN=</div>
          <div>
              NITF_ISDCDT=</div>
          <div>
              NITF_ISDCTP=</div>
          <div>
              NITF_ISDCXM=</div>
          <div>
              NITF_ISDG=</div>
          <div>
              NITF_ISDGDT=</div>
          <div>
              NITF_ISORCE=</div>
          <div>
              NITF_ISREL=</div>
          <div>
              NITF_ISSRDT=</div>
          <div>
              NITF_ONAME=</div>
          <div>
              NITF_OPHONE=</div>
          <div>
              NITF_OSTAID=CSA</div>
          <div>
              NITF_PJUST=R</div>
          <div>
              NITF_PVTYPE=INT</div>
          <div>
              NITF_STYPE=BF01</div>
          <span>  NITF_TGTID=</span><br>
        </span></div>
      <hr tabindex="-1">
      <div id="divRplyFwdMsg" dir="ltr"><b>From:</b> Even Rouault
        <a class="moz-txt-link-rfc2396E" href="mailto:even.rouault@spatialys.com"><even.rouault@spatialys.com></a><br>
        <b>Sent:</b> April 7, 2021 5:23 AM<br>
        <b>To:</b> ni hao <a class="moz-txt-link-rfc2396E" href="mailto:ni_hao88@hotmail.com"><ni_hao88@hotmail.com></a>;
        <a class="moz-txt-link-abbreviated" href="mailto:gdal-dev@lists.osgeo.org">gdal-dev@lists.osgeo.org</a> <a class="moz-txt-link-rfc2396E" href="mailto:gdal-dev@lists.osgeo.org"><gdal-dev@lists.osgeo.org></a><br>
        <b>Subject:</b> Re: [gdal-dev] NITF Int16 image size limit
        <div> </div>
      </div>
      <div>
        <p>Hi,</p>
        <p>Actually the error message doesn't look like there's a uint32
          overflow as it displays a file offset that is beyond the 4 GB
          limit.</p>
        <p>Did you check that your file size is at least 4324329780 +
          4194302 = 4 328 524 082 bytes ? I guess it is not, or then
          there's some mysterious error in the low level GDAL I/O layer.</p>
        <p>So if your file is smaller than that,</p>
        <p>- either it is corrupted (would be good if you could check
          that with a non-GDAL implementation)</p>
        <p>- or there's a bug in the NITF driver in how it computes
          offsets. Could you report the values of the NITF_ABPP, NITF_IC
          and NITF_IMODE metadata reported by gdalinfo ?<br>
        </p>
        <p>Even<br>
        </p>
        <div class="x_moz-cite-prefix">Le 07/04/2021 à 02:09, ni hao a
          écrit :<br>
        </div>
        <blockquote type="cite">
          <div><span>gdalinfo output i:</span><br>
          </div>
          <div><span>Driver: NITF/National Imagery Transmission Format
              <div>Files: ##CH_CV_GRD\imagery\12499.ntf</div>
              <div>Size is 61180, 19694</div>
              <div>Coordinate System is `'</div>
              <div>GCP Projection =</div>
              GEOGCS["WGS 84",   ... ]<br>
            </span></div>
          <div><span>Band 1 Block=1024x1024 Type=UInt16,
              ColorInterp=Gray</span></div>
          <div><span>Band 2 Block=1024x1024 Type=UInt16,
              ColorInterp=Gray</span><br>
          </div>
          <div><br>
          </div>
          <div>Also I am sure it is the UInt16 overflow, as the message
            is:</div>
          <div>ERROR 3: Unable to read 4194302 byte block from
            4324329780<br>
          </div>
          <div>and 2**32 = 4294967296</div>
          <div><br>
          </div>
          <hr tabindex="-1">
          <div id="x_divRplyFwdMsg" dir="ltr"><b>From:</b> <a
              class="x_moz-txt-link-abbreviated"
              href="mailto:bradh@frogmouth.net" moz-do-not-send="true">
              bradh@frogmouth.net</a> <a
              class="x_moz-txt-link-rfc2396E"
              href="mailto:bradh@frogmouth.net" moz-do-not-send="true">
              <bradh@frogmouth.net></a><br>
            <b>Sent:</b> April 6, 2021 8:17 PM<br>
            <b>To:</b> Even Rouault <a class="x_moz-txt-link-rfc2396E"
              href="mailto:even.rouault@spatialys.com"
              moz-do-not-send="true">
              <even.rouault@spatialys.com></a>; ni hao <a
              class="x_moz-txt-link-rfc2396E"
              href="mailto:ni_hao88@hotmail.com" moz-do-not-send="true">
              <ni_hao88@hotmail.com></a><br>
            <b>Cc:</b> <a class="x_moz-txt-link-abbreviated"
              href="mailto:gdal-dev@lists.osgeo.org"
              moz-do-not-send="true">
              gdal-dev@lists.osgeo.org</a> <a
              class="x_moz-txt-link-rfc2396E"
              href="mailto:gdal-dev@lists.osgeo.org"
              moz-do-not-send="true">
              <gdal-dev@lists.osgeo.org></a><br>
            <b>Subject:</b> Re: [gdal-dev] NITF Int16 image size limit
            <div> </div>
          </div>
          <div>
            <div dir="auto">Also if you can show the headers (e.g. from
              gdalinfo), that might provide indication of exactly which
              value is the issue.
              <div dir="auto"><br>
              </div>
              <div dir="auto">Brad</div>
            </div>
            <div class="x_x_gmail_extra"><br>
              <div class="x_x_gmail_quote">On 7 Apr. 2021 04:00, Even
                Rouault <a class="x_moz-txt-link-rfc2396E"
                  href="mailto:even.rouault@spatialys.com"
                  moz-do-not-send="true">
                  <even.rouault@spatialys.com></a> wrote:<br
                  type="attribution">
                <blockquote class="x_x_quote">
                  <div>
                    <p>Please keep the mailing list copied (I re-added
                      it)</p>
                    <p>This is not the most recent one (3.2.2 is), but
                      skimming quickly through release notes, I can't
                      see anything directly related, so it might still
                      be current with the recent versions. Altough I see
                      that a bunch of "avoid unsigned integer overflow"
                      type of fixes have been done since 2.4, so some of
                      them might have accidentally fixed the issue.</p>
                    <p>If you can try with the latest version ( Docker
                      images mentioned at <a
                        href="https://github.com/OSGeo/gdal/tree/master/gdal/docker"
                        moz-do-not-send="true">
https://github.com/OSGeo/gdal/tree/master/gdal/docker</a>, Conda, etc
                      .), that could be interesting.<br>
                    </p>
                    <p><br>
                    </p>
                    <div>Le 06/04/2021 à 19:51, ni hao a écrit :<br>
                    </div>
                    <blockquote>
                      <div>Hi Even,</div>
                      <div><br>
                      </div>
                      <div>The version is 2.4.4.</div>
                      <hr>
                      <div dir="ltr"><b>From:</b> Even Rouault <a
                          href="mailto:even.rouault@spatialys.com"
                          moz-do-not-send="true">
                          <even.rouault@spatialys.com></a><br>
                        <b>Sent:</b> April 6, 2021 2:46 PM<br>
                        <b>To:</b> ni hao <a
                          href="mailto:ni_hao88@hotmail.com"
                          moz-do-not-send="true"><ni_hao88@hotmail.com></a>;
                        <a href="mailto:gdal-dev@lists.osgeo.org"
                          moz-do-not-send="true">gdal-dev@lists.osgeo.org</a>
                        <a href="mailto:gdal-dev@lists.osgeo.org"
                          moz-do-not-send="true">
                          <gdal-dev@lists.osgeo.org></a><br>
                        <b>Subject:</b> Re: [gdal-dev] NITF Int16 image
                        size limit
                        <div> </div>
                      </div>
                      <div>
                        <p>Shawn,</p>
                        <p>This sounds more like a unwanted integer
                          overflow somewhere in the NITF driver.  Is
                          this with a recent GDAL ? If so, please file a
                          bug at
                          <a
                            href="https://github.com/OSGeo/gdal/issues/new"
                            moz-do-not-send="true">https://github.com/OSGeo/gdal/issues/new</a>
                          with all the details needed. A link to the
                          dataset would be ideal, but otherwise please
                          provide in the ticket description the output
                          of "gdalinfo your.nitf"</p>
                        <p>Even<br>
                        </p>
                        <div>Le 06/04/2021 à 19:35, ni hao a écrit :<br>
                        </div>
                        <blockquote>
                          <div>
                            <p>Hi list,</p>
                            <p><br>
                            </p>
                            <p>I encountered problem ingesting a
                              large NITF Int16 image with GDAL:</p>
                            <p>NITF 2-band image in Int16 format, 4.3
                              GB. It has 19690 lines x 61180 pixels. 
                               The upper 90% of the image looks fine. 
                               But the last 2000 lines fail to load by
                              GDAL.   That coincides with the 32-bit
                              boundary.  </p>
                            <p><br>
                            </p>
                            <p>Note that another larger NITF in 32-bit
                              float, 10.1 GB file loads fine.  </p>
                            <p><br>
                            </p>
                            <p>Is there a hard limit on NITF Int16
                              file size?</p>
                            <div>
                              <div>
                                <div>
                                  <div> </div>
                                  <div>Thanks,<br>
                                    Shawn</div>
                                </div>
                              </div>
                            </div>
                            <br>
                          </div>
                          <br>
                          <fieldset></fieldset>
                          <pre>_______________________________________________
gdal-dev mailing list
<a href="mailto:gdal-dev@lists.osgeo.org" moz-do-not-send="true">gdal-dev@lists.osgeo.org</a>
<a href="https://lists.osgeo.org/mailman/listinfo/gdal-dev" moz-do-not-send="true">https://lists.osgeo.org/mailman/listinfo/gdal-dev</a>
</pre>
                        </blockquote>
                        <pre>-- 
<a href="http://www.spatialys.com" moz-do-not-send="true">http://www.spatialys.com</a>
My software is free, but my time generally not.</pre>
                      </div>
                    </blockquote>
                    <pre>-- 
<a href="http://www.spatialys.com" moz-do-not-send="true">http://www.spatialys.com</a>
My software is free, but my time generally not.</pre>
                  </div>
                </blockquote>
              </div>
              <br>
            </div>
          </div>
          <br>
          <fieldset class="x_mimeAttachmentHeader"></fieldset>
          <pre class="x_moz-quote-pre">_______________________________________________
gdal-dev mailing list
<a class="x_moz-txt-link-abbreviated" href="mailto:gdal-dev@lists.osgeo.org" moz-do-not-send="true">gdal-dev@lists.osgeo.org</a>
<a class="x_moz-txt-link-freetext" href="https://lists.osgeo.org/mailman/listinfo/gdal-dev" moz-do-not-send="true">https://lists.osgeo.org/mailman/listinfo/gdal-dev</a>
</pre>
        </blockquote>
        <pre class="x_moz-signature" cols="72">-- 
<a class="x_moz-txt-link-freetext" href="http://www.spatialys.com" moz-do-not-send="true">http://www.spatialys.com</a>
My software is free, but my time generally not.</pre>
      </div>
    </blockquote>
    <pre class="moz-signature" cols="72">-- 
<a class="moz-txt-link-freetext" href="http://www.spatialys.com">http://www.spatialys.com</a>
My software is free, but my time generally not.</pre>
  </body>
</html>