<div dir="ltr"><div class="gmail_extra"><div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<div>
> I've got a UltraSPARC III machine with Solaris 10 that I just checked.<br>
<br>
</div>Still chugging along I bet.<br></blockquote><div><br></div><div>Some days yes, some days no. The PSU shorted out and filled the server room w/ smoke a few years ago. It was slated for the trash but it seemed like a shame to toss a neat old RISC box so I re-wired a scrap Dell PSU onto the connector of the fried Sun PSU and that's kept it going for the past 2 years. Far from reliable though.<br>
</div><div> </div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div>
> Try applying the attached patch if you're using SolarisStudio to fix the build problem.<br>
<br>
</div></blockquote><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">Ah yes .. that chestnut. I have seen this in gettext also<br></blockquote><div>
<br></div><div>It was a leftover from the cmakeify of liblzma. It's actually not a GCC-ism but a badly behaved suncc. It seems that suncc on Solaris intrinsically defines the type _Bool versus requiring stdbool.h. Interestingly, suncc on Linux x64 doesn't. The patch checks to see if it's intrinsically defined to prevent redefinition.<br>
<br><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">I have tons, literally, of weird hardware around me. Including a few<br>
from the late 90's that are still running and that includes a DEC AlphaServer<br>
4100 series. Not sure what to do with that.</blockquote><div><br></div><div>A shame. Alpha and PA-RISC have gone the way of the DoDo bird, i.e. long since EOL'd architectures. It's tough to justify even the power cost of keeping those running these days. <br>
<br></div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"> However Sparc servers both old and new I have. Well, newish anyways.<br>
</blockquote>
<div><br>Solaris 10 and 11 would be ideal, but whatever platform you can contribute would be great!<br><br></div><div> <blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
A big fat binary over 146MB in size. Staggering.<br></blockquote><div><br></div><div>The re-distributable binaries are so huge because of static linking of dependencies and included debug information. Dynamic linking and aggressive LTO with dead code elimination would shrink the size *considerably* but result in far less portable binaries, so not so suitable for redistribution.<br>
</div><br> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">I will say that I do have a cmake binary however its test results are<br>
not pretty:<br></blockquote><div><br></div><div>cmake --version will tell you what version you have. 2.8.4 or greater should work for the dashboard builds. If you use OpenCSW for 3rd party packages then you can install the CSWcmake package to get 2.8.12, easily new enough to drive the dashboard. If you want to build from source instead though, then you can see from last night's nightly dashboard:<br>
<br>Solaris 10 + SolarisStudio<br></div><a href="http://open.cdash.org/buildSummary.php?buildid=3470493" target="_blank">http://open.cdash.org/buildSummary.php?buildid=3470493</a><br>No build failures, 4 test failures<br>
<br>Solaris 10 + GCC (from OpenCSW)<br>
<a href="http://open.cdash.org/buildSummary.php?buildid=3470687" target="_blank">http://open.cdash.org/buildSummary.php?buildid=3470687</a><br></div><div>No build failures, 5 test failures<br><br></div><div>This is for the staging/next branch, which meant it should be in master soon. <br>
<br></div><div>As for setting up the nightly build, you should be able
to bootstrap and build the current master with gcc. Even if there's
still some test failures, you can use that cmake to drive the nightly
build.<br>
</div><br clear="all"><div><div dir="ltr">- Chuck<br></div></div>
<br></div></div>