This is the mail archive of the
mailing list for the GCC project.
Re: [PATCH] Fix PR middle-end/7847
> When I was looking into this bug a long time ago I believed that the
> issue was in the propagation of the attribute to the inner objects,
> but that obviously is not true.
I just looked into this again and realized that I was wrong when I was saying
that everything would be fine if the optimization didn't turn (mem:BLK) into
(mem:SI), because the same transformation is applied later, line 7315.
So the compiler still emits wrong code when the optimization is commented
On the other hand, the fix works with or without the optimization.