On Linux/x86-64, revision 172698 gave: g++ -O3 -funroll-loops -ffast-math -fwhole-program -flto=jobserver -fuse-linker-plugin .... ... In member function '_ZN8ggStringaSERKS_.constprop.760': lto1: internal compiler error: Segmentation fault Please submit a full bug report, with preprocessed source if appropriate. See <http://gcc.gnu.org/bugs.html> for instructions. make[4]: *** [/tmp/cc7AjsKg.ltrans23.ltrans.o] Error 1 lto-wrapper: make returned 2 exit status /usr/local/bin/ld: lto-wrapper failed collect2: ld returned 1 exit status specmake[3]: *** [eon] Error 1 Revision 172608 is OK.
It is caused by revision 172609: http://gcc.gnu.org/ml/gcc-cvs/2011-04/msg00804.html
I see the same for 483.xalancbmk: In member function '_ZN10xalanc_1_818XalanDOMStringPoolC2Emmm.constprop.17718': lto1: internal compiler error: Segmentation fault Please submit a full bug report, with preprocessed source if appropriate. See <http://gcc.gnu.org/bugs.html> for instructions. make[1]: *** [/tmp/ccfhgYqS.ltrans10.ltrans.o] Error 1 but eon may be nicer to analyze indeed.
The function _ZN10xalanc_1_818XalanDOMStringPoolC2Emmm is the one causing PR 48585 (caused by an earlier commit of mine). So these might actually be two different bugs (is is a segfault in the LTRANS stage, I suppose) ...or it is a better manifestation of the same bug, it is hard to tell. I guess I should update my source anyway.
Looking at our periodic tester results, it seems that it builds now and it started building at the same time 483.xalancbmk did. So I believe it indeed was a duplicate of PR 48585. *** This bug has been marked as a duplicate of bug 48585 ***