On Mon, Mar 26, 2012 at 8:51 PM, Steven Bosscher <stevenb....@gmail.com> wrote: > On Mon, Mar 26, 2012 at 8:22 PM, H.J. Lu <hjl.to...@gmail.com> wrote: >> It may have caused: >> >> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52730 > > It certainly seems so. Looking into it...
Java is the culprit. I'm going to commit the following patch to paper over the issue for now. Index: varasm.c =================================================================== --- varasm.c (revision 185823) +++ varasm.c (working copy) @@ -2170,6 +2170,10 @@ If it's not, we should not be calling this function. */ gcc_assert (asm_out_file); + /* In a perfect world, the following condition would be true. + Sadly, the Java and Go front ends emit assembly *from the front end*, + bypassing the call graph. See PR52739. Fix before GCC 4.8. */ +#if 0 /* This function should only be called if we are expanding, or have expanded, to RTL. Ideally, only final.c would be calling this function, but it is @@ -2177,6 +2181,7 @@ for further discussion. */ gcc_assert (cgraph_state == CGRAPH_STATE_EXPANSION || cgraph_state == CGRAPH_STATE_FINISHED); +#endif if (!DECL_P (decl) || !DECL_EXTERNAL (decl) || !TREE_PUBLIC (decl)) return;