https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
Bernd Edlinger changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #21 from Bernd Edlinger ---
Author: edlinger
Date: Sun Feb 22 19:38:53 2015
New Revision: 220899
URL: https://gcc.gnu.org/viewcvs?rev=220899&root=gcc&view=rev
Log:
2015-02-22 Bernd Edlinger
PR fortran/64980
PR for
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #20 from Bernd Edlinger ---
(In reply to Dominique d'Humieres from comment #12)
> In addition Bernd's patch fixes/hides the ICE for pr61960.
this pr is fixed by this hunk alone:
--- gcc/fortran/trans-expr.c(revision 220662)
+++ g
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #19 from Dominique d'Humieres ---
> BTW: should we add the original test case from pr64230 the test suite,
> because class_allocate_18.f90 failed to spot this regression?
Why not? Better safe than sorry. Note that the original test d
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #18 from Bernd Edlinger ---
(In reply to Dominique d'Humieres from comment #17)
> > I think this updated patch fixes all mentioned test cases.
>
> Confirmed, bootstrapped and regtested cleanly.
>
> > but there are many regressions,
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #17 from Dominique d'Humieres ---
> I think this updated patch fixes all mentioned test cases.
Confirmed, bootstrapped and regtested cleanly.
> but there are many regressions, which would need to be fixed before
> it makes sense to
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #16 from Bernd Edlinger ---
(In reply to Mikael Morin from comment #14)
> Bernd, do you have a reliable way to test a patch, checking for aliasing
> violations?
No, all I can do is run the test suite on my armv7l-unknown-linux-gnueab
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #15 from Bernd Edlinger ---
Created attachment 34758
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=34758&action=edit
Updated patch that also fixes pr64230.f90
(In reply to Dominique d'Humieres from comment #12)
> With any of t
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #14 from Mikael Morin ---
Bernd, do you have a reliable way to test a patch, checking for aliasing
violations?
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #13 from Mikael Morin ---
Created attachment 34757
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=34757&action=edit
Yet another attempt using TYPE_CANONICAL
this tries to remove any use of VIEW_CONVERT_EXPR by setting (well, tr
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #12 from Dominique d'Humieres ---
With any of the patches the first test for pr64230 fails at run time with
At line 35 of file pr64230.f90
Fortran runtime error: Attempt to DEALLOCATE unallocated 't'
but the test gfortran.dg/class_a
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #11 from Mikael Morin ---
Created attachment 34755
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=34755&action=edit
Another proposed fix
(In reply to Bernd Edlinger from comment #10)
> Yes, but I have no idea how to know how t
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #10 from Bernd Edlinger ---
(In reply to Mikael Morin from comment #9)
>
> (In reply to Bernd Edlinger from comment #8)
> > The call of get_d_position is now rewritten as follows:
> >
> > this.22 = VIEW_CONVERT_EXPR > __class_muli
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #9 from Mikael Morin ---
(In reply to Bernd Edlinger from comment #8)
> I will post this new patch, which combines Mikael's patch and
> fixes class_41.f03 and these test cases,
Actually, my patch was not supposed to be a real fix, ra
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #8 from Bernd Edlinger ---
Created attachment 34751
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=34751&action=edit
Proposed Fix
OK, now I see the original test case exposes an aliasing violation
when node is passed to node._v
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #7 from Dominique d'Humieres ---
> With this patch all fortran test cases pass on x86_64.
> And the ICE goes away in the reduced test example.
> Does it work for you?
It works for me for all the tests in this PR, without regression,
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
Bernd Edlinger changed:
What|Removed |Added
CC||bernd.edlinger at hotmail dot
de
--- C
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
--- Comment #5 from Mikael Morin ---
(In reply to Mikael Morin from comment #4)
> This cures it:
Another way to cure it:
Index: trans-expr.c
===
--- trans-expr.c(révision 22051
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64980
Richard Biener changed:
What|Removed |Added
Priority|P3 |P4
Summary|[5.0 Regression]
19 matches
Mail list logo
