https://sourceware.org/bugzilla/show_bug.cgi?id=32896
--- Comment #17 from Matthew Malcomson <mmalcomson at nvidia dot com> --- (In reply to H.J. Lu from comment #16) > (In reply to Matthew Malcomson from comment #14) > > (In reply to H.J. Lu from comment #12) > > > Created attachment 16054 [details] > > > A patch > > > > > > Please try this one on the current master branch. > > > > It seems to exhibit the same problem. > > > > I think it's doing the same as the previous patch -- in my artificial linker > > script I've re-ordered .tbss and .tdata, which means that the .tbss section > > is the first one seen. > > > > Please check the updated patch. This is at least working for the testcases that I've managed to produce so far. I am vaguely wondering if there's any non-TLS version of the original problem (NOBITS section determining the file offset of a segment) but have not been able to get a testcase for such a thing yet. Hope to have a little more of an attempt towards that tomorrow. -- You are receiving this mail because: You are on the CC list for the bug.
