[Bug tree-optimization/71170] [7 Regression] ICE in rewrite_expr_tree, at tree-ssa-reassoc.c:3898

["kugan at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71170

--- Comment #17 from kugan at gcc dot gnu.org ---
(In reply to kugan from comment #15)
> (In reply to David Binderman from comment #14)
> > (In reply to Jakub Jelinek from comment #12)
> > > Is it still broken?
> > 
> > I think so. Attachment seems to break svn 237286, dated today.
> 
> The issue with tree-reassoc is fixed now. Attached code ICE with
> -fno-tree-reassoc also. Here is reduced testcase that shows the current ICE.
> I think we should close this PR and create a new one for this.
> 
> cat b.c
> struct {
>   int error;
> } *a;
> 
> extern int fz_push_try ();
> int pdf_page_render() { return fz_push_try() && (a->error = __sigsetjmp()); }
> 
> ./gcc/cc1 -O2 b.c -fno-tree-reassoc
>  pdf_page_render
> b.c: In function âpdf_page_renderâ:
> b.c:6:61: warning: implicit declaration of function â__sigsetjmpâ
> [-Wimplicit-function-declaration]
>  int pdf_page_render() { return fz_push_try() && (a->error = __sigsetjmp());
> }
>                                                              ^~~~~~~~~~~
> 
> Analyzing compilation unit
> Performing interprocedural optimizations
>  <*free_lang_data> <visibility> <build_ssa_passes>b.c:6:1: error: definition
> in block 4 does not dominate use in block 5
>  int pdf_page_render() { return fz_push_try() && (a->error = __sigsetjmp());
> }
>  ^~~
> for SSA_NAME: a.1_2 in statement:
> # .MEM_14 = VDEF <.MEM_13>
> a.1_2->error = _3;
> b.c:6:1: internal compiler error: verify_ssa failed
> 0xdacfab verify_ssa(bool, bool)
> 	../../test/gcc/tree-ssa.c:1039
> 0xac3c87 execute_function_todo
> 	../../test/gcc/passes.c:1971
> 0xac2ce7 do_per_function
> 	../../test/gcc/passes.c:1648
> 0xac3e1c execute_todo
> 	../../test/gcc/passes.c:2016
> Please submit a full bug report,
> with preprocessed source if appropriate.
> Please include the complete backtrace with any bug report.
> See <http://gcc.gnu.org/bugs.html> for instructions.

This new issue is duplicate of PR71104 and started with r235817. This PR which
is about tree-reassoc is fixed and can be closed.