<html><head><meta http-equiv="Content-Type" content="text/html charset=iso-8859-1"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;">You need to fix two places:<div><br></div><div>Line 1256 and Line 1272. You only fixed one of them. </div><div><br><div><div>On Jun 17, 2013, at 11:50 AM, Pekka Jääskeläinen <<a href="mailto:pekka.jaaskelainen@tut.fi">pekka.jaaskelainen@tut.fi</a>> wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite"><div style="letter-spacing: normal; orphans: auto; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: auto; word-spacing: 0px; -webkit-text-stroke-width: 0px;">OK,<br><br>(Sorry for the wrong mailing list.)<br><br>Committed the fix. It seems to me that the only place where the address space<br>can get lost is when bitcasting the original scalar ptr to a vector one. Is this<br>right?<br><br>What shall be the protocol on getting fixes to the maintenance release?<br>Should I just commit the fix to the 3.3 branch as well (after LLVM 3.3<br>is actually out)?<br><br>On 06/17/2013 09:01 PM, Nadav Rotem wrote:<br><blockquote type="cite">Pekka,<br><br>LGTM. Please also make sure that we are not dropping the address space for Loads.<br><br>Thanks,<br>Nadav<br><br>On Jun 17, 2013, at 10:57 AM, Eli Friedman <<a href="mailto:eli.friedman@gmail.com">eli.friedman@gmail.com</a><br><<a href="mailto:eli.friedman@gmail.com">mailto:eli.friedman@gmail.com</a>>> wrote:<br><br><blockquote type="cite">Forwarding to the correct list.<br><br>---------- Forwarded message ----------<br>From:*Pekka Jääskeläinen*<<a href="mailto:pekka.jaaskelainen@tut.fi">pekka.jaaskelainen@tut.fi</a><br><<a href="mailto:pekka.jaaskelainen@tut.fi">mailto:pekka.jaaskelainen@tut.fi</a>>><br>Date: Mon, Jun 17, 2013 at 8:06 AM<br>Subject: [patch] fix loopvectorizer bug with multiple address spaces PR16306<br>To:cfe-<a href="mailto:commits@cs.uiuc.edu">commits@cs.uiuc.edu</a><span class="Apple-converted-space"> </span><<a href="mailto:cfe-commits@cs.uiuc.edu">mailto:cfe-commits@cs.uiuc.edu</a>><br><br><br>Hello,<br><br>Attached patch fixes a severe regression caused by LoopVectorizer when<br>vectorizing loops with memory accesses to non-zero address spaces. It<br>simply drops the AS info. It's a rather nasty problem as there is no<br>way (at least known to me) to disable the loop vectorizer *only* from<br>clang/opt command line.<br><br>Seems like an obvious fix.<br><br><a href="http://llvm.org/bugs/show_bug.__cgi?id=16306">http://llvm.org/bugs/show_bug.__cgi?id=16306</a><br><<a href="http://llvm.org/bugs/show_bug.cgi?id=16306">http://llvm.org/bugs/show_bug.cgi?id=16306</a>><br><br>The patch applies to trunk and 3.3 branches. Tom, could you apply<br>this to the (upcoming?) 3.3 maintenance branch as well?<br><br>Thanks,<br>--<br>--Pekka<br><br>_______________________________________________<br>cfe-commits mailing list<br><a href="mailto:cfe-commits@cs.uiuc.edu">cfe-commits@cs.uiuc.edu</a><span class="Apple-converted-space"> </span><<a href="mailto:cfe-commits@cs.uiuc.edu">mailto:cfe-commits@cs.uiuc.edu</a>><br><a href="http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits">http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits</a><br><br><br><loopvec-address-space-fix.patch><br></blockquote><br></blockquote><br><br>--<span class="Apple-converted-space"> </span><br>--PJ</div></blockquote></div><br></div></body></html>