[Bug middle-end/42068] [4.5 regression] ICE in function_and_variable_visibility with static common vars.
--- Comment #36 from rguenth at gcc dot gnu dot org 2010-03-08 11:59 --- (In reply to comment #35) > I find this bug mildly confusing. As the bootstrap / ada issues are fixed > (are they?) I'll split out static int a __attribute__((common)) ICE to > separate > PR. -> PR43288 -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068
[Bug middle-end/42068] [4.5 regression] ICE in function_and_variable_visibility with static common vars.
--- Comment #35 from rguenth at gcc dot gnu dot org 2010-03-08 11:56 --- I find this bug mildly confusing. As the bootstrap / ada issues are fixed (are they?) I'll split out static int a __attribute__((common)) ICE to separate PR. Please close this bug if the bootstra / ada issues are indeed fixed. They are P4 (non-primary/secondary arch, and Ada). -- rguenth at gcc dot gnu dot org changed: What|Removed |Added Priority|P1 |P4 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068
[Bug middle-end/42068] [4.5 regression] ICE in function_and_variable_visibility with static common vars.
--- Comment #34 from hubicka at gcc dot gnu dot org 2010-03-07 20:49 --- Hi, since this is blocker for a release and I can't reproduce the problem myself, if there any hope to get a backtrace? We can also just silence the sanity check for 4.5 for time being, but the proposed patch should solve the problem in proper way. Honza -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068
[Bug middle-end/42068] [4.5 regression] ICE in function_and_variable_visibility with static common vars.
--- Comment #33 from hubicka at ucw dot cz 2010-01-22 14:52 --- Subject: Re: [4.5 regression] ICE in function_and_variable_visibility with static common vars. > I'm not sure what Jan is asking for exactly, but in an attempt to be helpful > here is a traceback using the 20091124 sources for the reduced testcase on > alpha-openvms. The debugging info is messed up for me too, but its just the > source line correlation records numbers, the call stack looks reasonable. If I > can be of further assistance please ask. The backtrace should be definitely enough. What I need is trace of the setfault reported in comment #19. It should happen on mainline with the proposed patch (attached in the PR) minus changes to Ada directory (that already was commited to mainline in modified form) applied. There is still bug in handling user defined comdat functions, so I would like to debug rest of the patch for mainline. Thanks, Honza -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068
[Bug middle-end/42068] [4.5 regression] ICE in function_and_variable_visibility with static common vars.
--- Comment #32 from rupp at gnat dot com 2010-01-21 17:36 --- I'm not sure what Jan is asking for exactly, but in an attempt to be helpful here is a traceback using the 20091124 sources for the reduced testcase on alpha-openvms. The debugging info is messed up for me too, but its just the source line correlation records numbers, the call stack looks reasonable. If I can be of further assistance please ask. $ gdb ../../gnat1 (gdb) br fancy_abort (gdb) r -quiet -I. p.adb (gdb) bt #0 0x084eee38 in fancy_abort (file=0x8b079b7 "../../gcc-head-src/gcc/ipa.c", line=417, function=0x8b07ae0 "function_and_variable_visibility") at ../../gcc-head-src/gcc/diagnostic.c:155 #1 0x088c8a52 in function_and_variable_visibility (whole_program=0 '\0') at ../../gcc-head-src/gcc/cgraph.h:270 #2 0x088c8d91 in local_function_and_variable_visibility () at ../../gcc-head-src/gcc/cgraph.h:270 #3 0x08653c37 in execute_one_pass (pass=0x8b6b440) at ../../gcc-head-src/gcc/function.h:176 #4 0x08654352 in execute_ipa_pass_list (pass=0x8b6b440) at ../../gcc-head-src/gcc/function.h:176 #5 0x088a737e in ipa_passes () at ../../gcc-head-src/gcc/tree.h:182 #6 0x088a74f0 in cgraph_optimize () at ../../gcc-head-src/gcc/tree.h:182 #7 0x088a5d80 in cgraph_finalize_compilation_unit () at ../../gcc-head-src/gcc/tree.h:182 #8 0x0815c6d1 in gnat_write_global_declarations () at ../../gcc-head-src/gcc/tree.h:182 #9 0x0870e1bd in compile_file () at ../../gcc-head-src/gcc/toplev.h:198 #10 0x0870fe4a in do_compile () at ../../gcc-head-src/gcc/toplev.h:198 #11 0x0870ff0f in toplev_main (argc=4, argv=0xbfea58a4) at ../../gcc-head-src/gcc/toplev.h:198 #12 0x0847b33b in main (argc=4, argv=0xbfea58a4) at ../../gcc-head-src/gcc/main.c:35 -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068
[Bug middle-end/42068] [4.5 regression] ICE in function_and_variable_visibility with static common vars.
--- Comment #31 from ro at CeBiTec dot Uni-Bielefeld dot DE 2010-01-21 17:11 --- Subject: Re: [4.5 regression] ICE in function_and_variable_visibility with static common vars. > --- Comment #30 from hubicka at ucw dot cz 2010-01-21 16:57 --- > Subject: Re: [4.5 regression] ICE in function_and_variable_visibility with > static common vars. > >> >> Indeed it does: with that chance, a full bootstrap on alpha-dec-osf4.0f >> and alpha-dec-osf5.1b succeeded. > > Thanks for confirmation! I would however like to proceed with rest of > patch in order to fix the C testcase. Unfortunately I was not able to > reproduce the ICE on my machine neither at cross or by bootstrapping at > random architectures available to me. > Do you think you can gather at least some data about the ICE? (i.e. > function name or so). Looking through the patch I don't see any obvious > mistakes. It might be side effect of moving RTL generation of > assembling alias, but that is needed to avoid RTL from being produced > too early before visibilities are set. Unfortunately, I've so far been unsuccessful in getting any debugger to work properly on the gnat1 binary. I've tried gdb mainline, but there's the same error, and indeed gnat1 cannot properly be processed by either objdump or stdump (a native Tru64 UNIX tool to dump symbolic debug information). This seems to partially be a problem with the bootstrap compiler (gcc 4.4.2 in my case), though: the gnat1 from my last successful bootstap can at least be processed by stdump, though gdb 7.0 still SEGVs on it ;-( I could try another mainline bootstrap, though: maybe I'll be more lucky this time or be able what causes gdb to crash. Which patch exactly would you like me to test? Are the ipa.c and varasm.c changes here http://gcc.gnu.org/bugzilla/attachment.cgi?id=19623&action=view still up to date? Rainer -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068
[Bug middle-end/42068] [4.5 regression] ICE in function_and_variable_visibility with static common vars.
--- Comment #30 from hubicka at ucw dot cz 2010-01-21 16:57 --- Subject: Re: [4.5 regression] ICE in function_and_variable_visibility with static common vars. > > Indeed it does: with that chance, a full bootstrap on alpha-dec-osf4.0f > and alpha-dec-osf5.1b succeeded. Thanks for confirmation! I would however like to proceed with rest of patch in order to fix the C testcase. Unfortunately I was not able to reproduce the ICE on my machine neither at cross or by bootstrapping at random architectures available to me. Do you think you can gather at least some data about the ICE? (i.e. function name or so). Looking through the patch I don't see any obvious mistakes. It might be side effect of moving RTL generation of assembling alias, but that is needed to avoid RTL from being produced too early before visibilities are set. Honza -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068
[Bug middle-end/42068] [4.5 regression] ICE in function_and_variable_visibility with static common vars.
--- Comment #29 from ro at CeBiTec dot Uni-Bielefeld dot DE 2010-01-20 16:37 --- Subject: Re: [4.5 regression] ICE in function_and_variable_visibility breaks Ada bootstrap > --- Comment #20 from hubicka at ucw dot cz 2010-01-18 12:57 --- > Subject: Re: [4.5 regression] ICE in > function_and_variable_visibility breaks Ada bootstrap > >> This is really messy: maybe I'll have some more luck with a cross >> compiler. > Indeed it is. I will try, but I had problem building the cross because of > missing a.out headers. Can you, please, try the change to ada directory > itself? That alone should be enough to cure the bootstrap failure (not the C > version of testcase). So I am testing it now and intend to commit it ahead of > rest of changes. Indeed it does: with that chance, a full bootstrap on alpha-dec-osf4.0f and alpha-dec-osf5.1b succeeded. Thanks Rainer -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068
[Bug middle-end/42068] [4.5 regression] ICE in function_and_variable_visibility with static common vars.
--- Comment #28 from simon at pushface dot org 2010-01-19 12:36 --- (In reply to comment #27) I am sorry for spreading confusion & wasting time, I had a different PR-resolution process in mind. Now that the reduced patch is committed to mainline, there is no longer an ICE during build of i86_64-apple-darwin10.2, powerpc-wrs-vxworks. So Richard is right in comment #23. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068
[Bug middle-end/42068] [4.5 regression] ICE in function_and_variable_visibility with static common vars.
--- Comment #27 from hubicka at ucw dot cz 2010-01-19 10:48 --- Subject: Re: [4.5 regression] ICE in function_and_variable_visibility with static common vars. > The problem is that the ICE-on-valid occurs while building the Ada RTS, and > that is a bootstrap issue for Ada (what I mean is, a build configured with > "--enable-languages=foo,ada" will fail). Does it happen also with the reduced version of the patch I comitted to mainline? I can't imagine how this one can trigger ICE, but it might be independent change in meantime as I was slow on fixing this bug :( Does the reduced testcase also trigger it? (I will try it on cross, since without useful backtrace I can't do much). I proofread the other part of patch but don't see anything obviously wrong here. Honza -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068
[Bug middle-end/42068] [4.5 regression] ICE in function_and_variable_visibility with static common vars.
--- Comment #26 from simon at pushface dot org 2010-01-18 20:54 --- (In reply to comment #22) > No longer bootstrap issue, but still ICE on valid. The problem is that the ICE-on-valid occurs while building the Ada RTS, and that is a bootstrap issue for Ada (what I mean is, a build configured with "--enable-languages=foo,ada" will fail). -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068
[Bug middle-end/42068] [4.5 regression] ICE in function_and_variable_visibility with static common vars.
--- Comment #25 from simon at pushface dot org 2010-01-18 20:41 --- OK on i86_64-apple-darwin10.2, powerpc-wrs-vxworks (hosted on i366-apple-darwin10.2). -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068
[Bug middle-end/42068] [4.5 regression] ICE in function_and_variable_visibility with static common vars.
--- Comment #24 from hubicka at gcc dot gnu dot org 2010-01-18 17:19 --- Subject: Bug 42068 Author: hubicka Date: Mon Jan 18 17:19:13 2010 New Revision: 156016 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=156016 Log: PR middle-end/42068 * gcc-interface/utils.c (create_var_decl_1): Do not set COMMON flag for unit local variables. Modified: trunk/gcc/ada/ChangeLog -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068
[Bug middle-end/42068] [4.5 regression] ICE in function_and_variable_visibility with static common vars.
--- Comment #23 from rguenth at gcc dot gnu dot org 2010-01-18 15:48 --- If it's now middle-end then we need to adjust the priority. -- rguenth at gcc dot gnu dot org changed: What|Removed |Added Keywords|build |ice-on-valid-code Priority|P4 |P1 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068
[Bug middle-end/42068] [4.5 regression] ICE in function_and_variable_visibility with static common vars.
--- Comment #22 from hubicka at gcc dot gnu dot org 2010-01-18 15:47 --- No longer bootstrap issue, but still ICE on valid. -- hubicka at gcc dot gnu dot org changed: What|Removed |Added AssignedTo|unassigned at gcc dot gnu |hubicka at gcc dot gnu dot |dot org |org Status|NEW |ASSIGNED Last reconfirmed|2010-01-06 22:48:22 |2010-01-18 15:47:12 date|| Summary|[4.5 regression] ICE in |[4.5 regression] ICE in |function_and_variable_visibi|function_and_variable_visibi |lity breaks Ada bootstrap |lity with static common ||vars. http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42068
