[Bug target/42818] Static C++ linking breakage "undefined reference to ___real__Znwj" and others in libcygwin.a(_cygwin_crt0_common.o)
--- Comment #1 from davek at gcc dot gnu dot org 2010-01-21 03:20 --- Test results: clean h...@155967, default test configuration using shared C++ linking: http://gcc.gnu.org/ml/gcc-testresults/2010-01/msg01628.html clean h...@155967, tested using static C++ linking: http://gcc.gnu.org/ml/gcc-testresults/2010-01/msg01900.html patched h...@155967, tested with static linking and the simple stage-3-safe fix: http://gcc.gnu.org/ml/gcc-testresults/2010-01/msg01901.html -- davek at gcc dot gnu dot org changed: What|Removed |Added Status|UNCONFIRMED |ASSIGNED Ever Confirmed|0 |1 Last reconfirmed|-00-00 00:00:00 |2010-01-21 03:20:38 date|| http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42818
[Bug target/42818] Static C++ linking breakage "undefined reference to ___real__Znwj" and others in libcygwin.a(_cygwin_crt0_common.o)
--- Comment #2 from davek at gcc dot gnu dot org 2010-01-21 03:57 --- Here's what's going on: - To implement operator new/delete replacement, we use ld and --wrap to redirect all references to the operator functions into redirection stubs in the cygwin dll. - Each DLL or executable records a list of which (if any) of those operators it provides replacements for. It does this with an array of function pointers in a part of the CRT pre-main() startup. In the array are weak references to all the replaceable operators; any for which overrides are visible at final-link time will get filled out with the addresses of those overrides, and the weakness of the references means they won't cause any functions that aren't already present in the final link to be pulled in from libraries. The references assign asm names to the symbols that use the "__real__" prefix in order that they don't see the Cygwin DLL wrapper functions, just the real actual implementations present. - As an optimisation, we don't apply the wrapper redirection when linking statically. That's because there's no need; any function replacement overrides present in the final link will just automatically override the libstdc++ versions by preventing them from ever being pulled out of the static libstdc++ archive at final-link time. That's all fine as it stands, but:- - Some recent optimisation of EH generation had the effect that, where previously every C++ object was always emitted with a reference to the gxx eh personality function, this reference is now only emitted when strictly needed. - This means that for the first time it became possible that the only references to libstdc++ in a C++ executable undergoing final link might be references to the operator new/delete function(s). - Since these are all redirected to the wrappers in the Cygwin DLL, that can mean there are no unresolved references to pull in anything at all from the libstdc++ import library, and so the resulting executable does not depend on nor cause the libstdc++ dll to be loaded into memory, meaning that the real versions of the replacement functions that the redirect stubs need to redirect to aren't present! - So, the way to prevent that is by also emitting an extern reference to the real version of the operator function in question, which causes the DLL to be imported (even though the imports aren't used directly but only via the redirection stubs in the DLL; the important point being that you have to load libstdc++ so that it can register the default versions of all the operators using its own array of weak (but fully-filled-out) function pointers. This is done in gcc/config/i386/winnt.c, in i386_pe_file_end(). - And that's where it goes wrong. The compiler doesn't know at the time it's emitting assembly whether the final object will be statically linked or not, so it really has no choice but to emit the reference; but if it does end up being statically linked, the wrappers aren't in effect, so there is no "__real__*" function. - That wouldn't be a problem: the toolchain doesn't care about the undefined symbol, because no relocs reference it, and the runtime loader doesn't give a stuff about symbols at all, caring only about the import and export tables. - Except for one final gotcha. Because the undefined references that are emitted to cause the DLL to be loaded are strong undefs - they have to be, since they have to cause an import stub to be pulled in from a library archive. So even though there are no relocs in the object that contains the strong undef, it has the knock-on effect that the weak references in the array of function pointers become strong. And that's where the link failure arises. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42818
[Bug target/42818] Static C++ linking breakage "undefined reference to ___real__Znwj" and others in libcygwin.a(_cygwin_crt0_common.o)
--- Comment #3 from davek at gcc dot gnu dot org 2010-01-21 04:00 --- Created an attachment (id=19670) --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19670&action=view) Simple stage-3-safe fix This is the simple fix that is safe for stage 3. By trivially tweaking the specs, we remove the don't-redirect-via-wrappers-if-statically-linking optimisation. This is a minor performance hit, but correctness combines with the need for low-risk changes in stage 3 to win out here. I'll talk about the more optimal fix (and the complications that mean it has to wait for next stage 1) next. As I've already run all the tests, and as I've just been appointed a maintainer, I'll check this in shortly :-) -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42818
[Bug target/42818] Static C++ linking breakage "undefined reference to ___real__Znwj" and others in libcygwin.a(_cygwin_crt0_common.o)
--- Comment #4 from davek at gcc dot gnu dot org 2010-01-21 04:56 --- Subject: Bug 42818 Author: davek Date: Thu Jan 21 04:56:38 2010 New Revision: 156105 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=156105 Log: PR target/42818 * config/i386/cygwin.h (CXX_WRAP_SPEC_LIST): Always apply wrappers, even when linking statically, for now. Modified: trunk/gcc/ChangeLog trunk/gcc/config/i386/cygwin.h -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42818
[Bug target/42818] Static C++ linking breakage "undefined reference to ___real__Znwj" and others in libcygwin.a(_cygwin_crt0_common.o)
--- Comment #5 from davek at gcc dot gnu dot org 2010-01-21 05:00 --- Fixed for 4.5.0, so setting milestone. Then I'll add a comment about the full fix and suspend it. BZ maintainers, apologies if this is the wrong thing to do; should I close it and reopen it? I'm not sure, whatever you're happy with is fine by me. -- davek at gcc dot gnu dot org changed: What|Removed |Added Target Milestone|--- |4.5.0 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42818
[Bug target/42818] Static C++ linking breakage "undefined reference to ___real__Znwj" and others in libcygwin.a(_cygwin_crt0_common.o)
--- Comment #6 from davek at gcc dot gnu dot org 2010-01-21 05:00 --- So, the optimal solution would be to avoid using the redirection wrappers when we're statically linking. That's still possible: those undefined references without relocs don't cause problems in and of themselves, but only because they turn the weak undefs that *do* have relocs strong. The obvious solution, since we're not using the redirection wrapper functions, is to also not include the array of weak function pointers. That would be trival to do by building two different versions of the CRT startup code in Cygwin itself, one for shared linking that contains the array, one for static linking that doesn't, and selecting the appropriate one at link time using a spec in the backend. That would work fine for --static, but as things stand now, it will still fail when just --static-libstdc++ is in use. This is because of the situation described in the two dependency PRs (Bug 41594 and Bug 41596): you can't test for --static-libstdc++ in a spec, because the code in g++spec.c removes it from the command-line. So, I'm going to leave this bug open and suspend it now I've committed the simple fix. Then, next stage 1, we'll get the situation with --static-libstdc++ resolved one way or another, and by that time I'll have added the alternative startup CRT object to the Cygwin runtime, and then I'll revert this fix and adjust the linker spec to select the correct startup variant. -- davek at gcc dot gnu dot org changed: What|Removed |Added Status|ASSIGNED|SUSPENDED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42818
[Bug target/42818] Static C++ linking breakage "undefined reference to ___real__Znwj" and others in libcygwin.a(_cygwin_crt0_common.o)
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42818 Andrew Pinski changed: What|Removed |Added Target Milestone|4.5.3 |---
[Bug target/42818] Static C++ linking breakage "undefined reference to ___real__Znwj" and others in libcygwin.a(_cygwin_crt0_common.o)
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42818 Richard Guenther changed: What|Removed |Added Target Milestone|4.5.2 |4.5.3 --- Comment #9 from Richard Guenther 2010-12-16 13:02:57 UTC --- GCC 4.5.2 is being released, adjusting target milestone.
[Bug target/42818] Static C++ linking breakage "undefined reference to ___real__Znwj" and others in libcygwin.a(_cygwin_crt0_common.o)
--- Comment #8 from rguenth at gcc dot gnu dot org 2010-07-31 09:29 --- GCC 4.5.1 is being released, adjusting target milestone. -- rguenth at gcc dot gnu dot org changed: What|Removed |Added Target Milestone|4.5.1 |4.5.2 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42818
[Bug target/42818] Static C++ linking breakage "undefined reference to ___real__Znwj" and others in libcygwin.a(_cygwin_crt0_common.o)
--- Comment #7 from rguenth at gcc dot gnu dot org 2010-04-06 11:20 --- GCC 4.5.0 is being released. Deferring to 4.5.1. -- rguenth at gcc dot gnu dot org changed: What|Removed |Added Target Milestone|4.5.0 |4.5.1 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42818
[Bug target/42818] Static C++ linking breakage "undefined reference to ___real__Znwj" and others in libcygwin.a(_cygwin_crt0_common.o)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=42818 Bug 42818 depends on bug 41596, which changed state. Bug 41596 Summary: handling of library-related options by g++spec.c causes a failure when generating pch. https://gcc.gnu.org/bugzilla/show_bug.cgi?id=41596 What|Removed |Added Status|NEW |RESOLVED Resolution|--- |FIXED
[Bug target/42818] Static C++ linking breakage "undefined reference to ___real__Znwj" and others in libcygwin.a(_cygwin_crt0_common.o)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=42818 Eric Gallager changed: What|Removed |Added CC||egallager at gcc dot gnu.org --- Comment #10 from Eric Gallager --- (In reply to Dave Korn from comment #6) > That would work fine for --static, but as things stand now, it will still > fail when just --static-libstdc++ is in use. This is because of the > situation described in the two dependency PRs (Bug 41594 and Bug 41596) Both bugs upon which this one depends have been closed; time to revise this one's SUSPENDED status?
