http://llvm.org/bugs/show_bug.cgi?id=15965

David Fang <[email protected]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED

--- Comment #9 from David Fang <[email protected]> ---
(In reply to comment #8)
> (In reply to comment #6)
> > Rafael, do you think any of your recent ABI/alignment/DataLayout patches may
> > have inadvertently fixed this issue along
> > http://llvm.org/bugs/show_bug.cgi?id=15901 ?
> > Could misalignment have caused what looked like memory corruption errors?
> > I'd like to identify which patches should be backported to the
> > powerpc-darwin8-rel-3.4 branch.  (If it was indeed your patches, then a HUGE
> > THANK YOU for a significant unblocking of stage-2.)
> 
> Maybe. Possible suspects are 197694, 197574 and 197572.

r196987 +r197574 +r197572 seems to have done it!

ADTTests works in stage-2 on the powerpc-darwin8-rel-3.4 branch.  I suspect
that TableGen in 15901 will probably work too.
r196987 was just a prerequisite for the latter to apply cleanly.  

Thank goodness, I didn't have to bisect.  It's painful here because
master/trunk doesn't build as-is -- I'd have to repeatedly rebase or merge my
patchset in the bisection process (and builds are slow).

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
LLVMbugs mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs

Reply via email to